APS Software vs. ERP: What's the Difference?

1
00:00:00,000 --> 00:00:02,480
Here's the problem, most manufacturers don't talk about.

2
00:00:02,480 --> 00:00:04,720
Alayed order lands, a machine stops,

3
00:00:04,720 --> 00:00:07,080
and material that looked available is actually sitting

4
00:00:07,080 --> 00:00:09,360
in the inspection waiting for a QC stamp.

5
00:00:09,360 --> 00:00:11,920
Meanwhile, someone's about to promise a customer a date

6
00:00:11,920 --> 00:00:13,640
the factory can't possibly meet,

7
00:00:13,640 --> 00:00:15,800
and the planner needs an answer before that happens.

8
00:00:15,800 --> 00:00:17,360
The ERP plan still exists,

9
00:00:17,360 --> 00:00:19,320
the sales order, the work order, the planned dates

10
00:00:19,320 --> 00:00:22,520
are all still there, but the physical conditions have changed,

11
00:00:22,520 --> 00:00:25,680
and a plan on paper doesn't clear a queue or free a machine,

12
00:00:25,680 --> 00:00:27,360
and that's where the split becomes clear.

13
00:00:27,360 --> 00:00:29,840
ERP records and controls the business transaction.

14
00:00:29,840 --> 00:00:32,560
APS, advance planning and scheduling,

15
00:00:32,560 --> 00:00:34,920
tests what can actually run under the conditions

16
00:00:34,920 --> 00:00:36,440
the factory is facing right now.

17
00:00:36,440 --> 00:00:38,360
Let's follow one order through both systems

18
00:00:38,360 --> 00:00:41,480
because the same order can tell two very different stories.

19
00:00:41,480 --> 00:00:44,080
One order, two different views of the factory,

20
00:00:44,080 --> 00:00:46,800
picture a hypothetical plan that produces several product

21
00:00:46,800 --> 00:00:47,600
families.

22
00:00:47,600 --> 00:00:49,640
Some jobs need the same machining center,

23
00:00:49,640 --> 00:00:52,560
others need a shared test station near the end of the route.

24
00:00:52,560 --> 00:00:54,800
Labor is finite, and only a few operators

25
00:00:54,800 --> 00:00:57,080
hold the right qualification for certain work.

26
00:00:57,080 --> 00:00:59,560
Then a customer calls and needs an order brought forward

27
00:00:59,560 --> 00:01:01,880
on the ERP side, that order looks familiar.

28
00:01:01,880 --> 00:01:04,040
You've got a sales order with a requested date,

29
00:01:04,040 --> 00:01:06,640
a bill of materials that lists the parts and raw materials,

30
00:01:06,640 --> 00:01:09,200
and a rooting with the approved production steps.

31
00:01:09,200 --> 00:01:12,080
Inventory records show what the system thinks is in stock,

32
00:01:12,080 --> 00:01:14,480
and a work order tells production what to build.

33
00:01:14,480 --> 00:01:15,520
All of that matters.

34
00:01:15,520 --> 00:01:17,560
Without it, you don't know what the customer bought,

35
00:01:17,560 --> 00:01:20,720
what to build, or what materials should move through the business.

36
00:01:20,720 --> 00:01:24,040
ERP gives the order its commercial and transactional identity,

37
00:01:24,040 --> 00:01:27,040
but the order still needs to pass through a real factory.

38
00:01:27,040 --> 00:01:29,120
The moment the planner asks whether it can move forward

39
00:01:29,120 --> 00:01:31,840
the questions change, which operation comes first,

40
00:01:31,840 --> 00:01:33,720
which machine can perform it.

41
00:01:33,720 --> 00:01:36,560
Whether that machine is available during the required shift,

42
00:01:36,560 --> 00:01:39,040
if another order already occupies that time,

43
00:01:39,040 --> 00:01:41,040
and whether an alternate machine can do the work

44
00:01:41,040 --> 00:01:43,320
with a different tool or operator.

45
00:01:43,320 --> 00:01:45,080
This is where APS looks at the same order

46
00:01:45,080 --> 00:01:45,920
through a different lens.

47
00:01:45,920 --> 00:01:48,560
It doesn't just see a routing as a sequence of name steps.

48
00:01:48,560 --> 00:01:50,520
It looks at each operation in time,

49
00:01:50,520 --> 00:01:53,000
checks the current capacity of the machines involved,

50
00:01:53,000 --> 00:01:55,640
considers the sequence of work already planned on them,

51
00:01:55,640 --> 00:01:57,960
and accounts for rules around product families,

52
00:01:57,960 --> 00:02:01,360
two limits, labor skills, and material timing.

53
00:02:01,360 --> 00:02:04,640
That last part sounds simple until you get into a real planning meeting.

54
00:02:04,640 --> 00:02:08,160
ERP may show material on hand, but the planner needs to know

55
00:02:08,160 --> 00:02:10,560
whether it will actually be ready when the operation needs it.

56
00:02:10,560 --> 00:02:13,240
Maybe it arrives at receiving, but still needs a quality check.

57
00:02:13,240 --> 00:02:14,960
Maybe another order consumes it first.

58
00:02:14,960 --> 00:02:16,440
Maybe it's sitting in the wrong location.

59
00:02:16,440 --> 00:02:17,680
Inventory is a business fact.

60
00:02:17,680 --> 00:02:19,720
Material readiness is a scheduling condition.

61
00:02:19,720 --> 00:02:22,520
The same goes for labor, a work center can show available hours,

62
00:02:22,520 --> 00:02:24,240
but that doesn't mean the right person is there

63
00:02:24,240 --> 00:02:25,720
when the job reaches that step.

64
00:02:25,720 --> 00:02:27,760
If the process needs a certified welder,

65
00:02:27,760 --> 00:02:29,960
general labor capacity doesn't solve it.

66
00:02:29,960 --> 00:02:32,880
The job either waits, moves to another qualified resource,

67
00:02:32,880 --> 00:02:34,160
or the schedule changes.

68
00:02:34,160 --> 00:02:36,040
Now think about that urgent customer request.

69
00:02:36,040 --> 00:02:38,800
The ERP question might be whether to release the work order

70
00:02:38,800 --> 00:02:41,040
or change its due date and priority.

71
00:02:41,040 --> 00:02:42,120
Those are valid.

72
00:02:42,120 --> 00:02:44,240
They control what enters the planning process

73
00:02:44,240 --> 00:02:45,880
and what the business commits to.

74
00:02:45,880 --> 00:02:48,120
But APS asks something more physical.

75
00:02:48,120 --> 00:02:51,320
If we push this order forward, what can we still complete on time?

76
00:02:51,320 --> 00:02:53,080
That means it has to test consequences.

77
00:02:53,080 --> 00:02:55,720
Maybe the urgent order can run next on machine 12,

78
00:02:55,720 --> 00:02:57,560
but only if a current job moves.

79
00:02:57,560 --> 00:03:00,160
And that current job might then miss the test station window

80
00:03:00,160 --> 00:03:01,320
later in the week.

81
00:03:01,320 --> 00:03:03,760
Or the urgent job might fit on an alternate machine,

82
00:03:03,760 --> 00:03:06,760
but only after a tool change that delays two smaller orders.

83
00:03:06,760 --> 00:03:08,240
The factory doesn't give you a free slot

84
00:03:08,240 --> 00:03:09,840
just because a customer needs one.

85
00:03:09,840 --> 00:03:11,440
This is why a planner can feel trapped

86
00:03:11,440 --> 00:03:14,240
between an approved plan and actual constraints.

87
00:03:14,240 --> 00:03:17,280
The approved plan may show the urgent order finishing on Friday,

88
00:03:17,280 --> 00:03:19,720
but the machine calendar tells a different story.

89
00:03:19,720 --> 00:03:22,280
The supervisor knows one operator will be absent

90
00:03:22,280 --> 00:03:24,760
and the material team knows a component is late.

91
00:03:24,760 --> 00:03:26,760
Each person holds part of the answer

92
00:03:26,760 --> 00:03:29,440
and the planner ends up joining those answers manually,

93
00:03:29,440 --> 00:03:30,880
often under time pressure.

94
00:03:30,880 --> 00:03:32,640
That's not a failure of the planner.

95
00:03:32,640 --> 00:03:34,520
It's what happens when planning records

96
00:03:34,520 --> 00:03:37,160
and planning decisions sit in the same mental bucket.

97
00:03:37,160 --> 00:03:39,640
ERP gives you the records that govern the order,

98
00:03:39,640 --> 00:03:42,600
demand, materials, routing, inventory,

99
00:03:42,600 --> 00:03:44,080
and the work order itself.

100
00:03:44,080 --> 00:03:47,160
APS takes those records and asks whether the current physical limits

101
00:03:47,160 --> 00:03:48,360
actually permit the plan.

102
00:03:48,360 --> 00:03:51,120
Both views need to exist, but they answer different questions.

103
00:03:51,120 --> 00:03:54,000
Confusion starts when a planned ERP date gets treated as proof

104
00:03:54,000 --> 00:03:56,200
that the factory has the capacity to deliver.

105
00:03:56,200 --> 00:03:58,120
What ERP actually does well?

106
00:03:58,120 --> 00:04:00,760
Before I get into where ERP planning hits its limit,

107
00:04:00,760 --> 00:04:02,080
let me give credit where it's due.

108
00:04:02,080 --> 00:04:04,360
ERP does one thing extremely well.

109
00:04:04,360 --> 00:04:07,000
It gives the business a single controlled source of truth

110
00:04:07,000 --> 00:04:11,280
for the facts that govern demand, supply, money, and accountability.

111
00:04:11,280 --> 00:04:12,400
Here's how it plays out.

112
00:04:12,400 --> 00:04:14,200
A customer order enters the system,

113
00:04:14,200 --> 00:04:15,840
a buyer raises a purchase order,

114
00:04:15,840 --> 00:04:17,520
warehouse staff posts a goods receipt,

115
00:04:17,520 --> 00:04:19,200
production confirms a quantity,

116
00:04:19,200 --> 00:04:21,080
finance needs to understand the cost.

117
00:04:21,080 --> 00:04:24,160
These aren't side tasks that happen around manufacturing.

118
00:04:24,160 --> 00:04:25,560
They're the core business process

119
00:04:25,560 --> 00:04:27,840
that lets the manufacturer buy, build, ship, invoice,

120
00:04:27,840 --> 00:04:29,440
and trace everything that happened.

121
00:04:29,440 --> 00:04:32,920
ERP keeps those transactions connected in a way nothing else does.

122
00:04:32,920 --> 00:04:34,240
Let's start with demand.

123
00:04:34,240 --> 00:04:38,000
The ERP system holds forecasts, customer orders, requested dates,

124
00:04:38,000 --> 00:04:41,440
and often the rules that determine how that demand should be supplied.

125
00:04:41,440 --> 00:04:44,200
It connects demand to inventory, open purchase orders,

126
00:04:44,200 --> 00:04:46,520
planned orders, and production orders.

127
00:04:46,520 --> 00:04:48,960
When sales asks whether an order exists,

128
00:04:48,960 --> 00:04:50,880
when purchasing needs to know what to buy,

129
00:04:50,880 --> 00:04:53,880
or when finance wants to know what stock the company owns,

130
00:04:53,880 --> 00:04:55,920
ERP should give the governed answer.

131
00:04:55,920 --> 00:04:57,360
That's its job.

132
00:04:57,360 --> 00:04:58,760
That system of record matters

133
00:04:58,760 --> 00:05:02,400
because nobody wants each department holding its own version of an order.

134
00:05:02,400 --> 00:05:04,840
If the planner changes a date in one local tool,

135
00:05:04,840 --> 00:05:08,240
purchasing can't quite keep working from a different date in another tool.

136
00:05:08,240 --> 00:05:10,920
ERP forces everyone to work from the same page.

137
00:05:10,920 --> 00:05:13,960
And it gives you a controlled transaction trail to prove what happened.

138
00:05:13,960 --> 00:05:16,840
Material requirements planning, MRP for short,

139
00:05:16,840 --> 00:05:18,400
sits right in the middle of this.

140
00:05:18,400 --> 00:05:20,840
MRP takes demand for finished goods

141
00:05:20,840 --> 00:05:23,280
and works backward through the bill of materials.

142
00:05:23,280 --> 00:05:25,600
If you need to produce 100 units of a product,

143
00:05:25,600 --> 00:05:28,680
MRP calculates the components and raw materials required,

144
00:05:28,680 --> 00:05:32,320
compares those needs with supply already on hand, or already ordered,

145
00:05:32,320 --> 00:05:35,520
and creates planned supply signals wherever shortages appear.

146
00:05:35,520 --> 00:05:37,720
That's a practical and genuinely useful job.

147
00:05:37,720 --> 00:05:40,920
A planned order tells production that more supply needs to exist.

148
00:05:40,920 --> 00:05:44,200
A purchase requisition tells purchasing that material needs buying.

149
00:05:44,200 --> 00:05:47,200
MRP doesn't require a planner to manually inspect every bolt,

150
00:05:47,200 --> 00:05:50,400
subassembly, and packaging item across a large product structure.

151
00:05:50,400 --> 00:05:53,520
It creates a structured view of demand and supply across time,

152
00:05:53,520 --> 00:05:55,000
and that saves real hours.

153
00:05:55,000 --> 00:05:57,600
But here's the thing, MRP depends on master data.

154
00:05:57,600 --> 00:06:00,960
And master data is where the quiet, unglamerous work lives.

155
00:06:00,960 --> 00:06:02,320
Every item needs a definition.

156
00:06:02,320 --> 00:06:05,280
A bill of materials needs to describe what goes into the product.

157
00:06:05,280 --> 00:06:08,360
A routing needs to describe the approved production path.

158
00:06:08,360 --> 00:06:11,520
Work centers need to represent where work can actually happen.

159
00:06:11,520 --> 00:06:13,680
Lead times need to give the system a working assumption

160
00:06:13,680 --> 00:06:16,880
about how long purchasing, production, and transfer activities take.

161
00:06:16,880 --> 00:06:18,240
None of this feels glamorous.

162
00:06:18,240 --> 00:06:20,480
It's still the foundation everything rests on.

163
00:06:20,480 --> 00:06:23,280
If the item data is wrong, purchasing plans the wrong supply.

164
00:06:23,280 --> 00:06:24,960
If the bill of materials is out of date,

165
00:06:24,960 --> 00:06:27,520
production consumes parts nobody expected.

166
00:06:27,520 --> 00:06:29,000
If the routing misses an operation,

167
00:06:29,000 --> 00:06:32,000
cost and planning both drift away from the physical process.

168
00:06:32,000 --> 00:06:34,680
ERP creates discipline because it forces the business

169
00:06:34,680 --> 00:06:36,960
to define these things in a common structure.

170
00:06:36,960 --> 00:06:39,560
And that structure is what makes everything else possible.

171
00:06:39,560 --> 00:06:40,920
It also handles control.

172
00:06:40,920 --> 00:06:43,920
A purchase order creates an approved commitment to a supplier.

173
00:06:43,920 --> 00:06:45,880
A production order authorizes work.

174
00:06:45,880 --> 00:06:49,800
A goods movement records stock entering, leaving or moving within the business.

175
00:06:49,800 --> 00:06:53,880
A confirmation records completed work, consumed material, or reported time,

176
00:06:53,880 --> 00:06:55,880
depending on how you set up the process.

177
00:06:55,880 --> 00:06:58,120
That transaction control supports traceability,

178
00:06:58,120 --> 00:07:01,360
inventory accuracy, costing, and financial close.

179
00:07:01,360 --> 00:07:04,680
For regulated industries, it can support lot and batch records too.

180
00:07:04,680 --> 00:07:07,960
For any manufacturer, it gives people a way to answer basic questions

181
00:07:07,960 --> 00:07:11,840
without relying only on memory or a spreadsheet someone updates after hours.

182
00:07:11,840 --> 00:07:14,160
So what assumptions does ERP actually plan with?

183
00:07:14,160 --> 00:07:18,160
Lead times represent expected durations, calendars define normal working periods.

184
00:07:18,160 --> 00:07:22,040
Capacity gets expressed in broader time buckets, days or weeks,

185
00:07:22,040 --> 00:07:25,080
rather than a minute by minute sequence on a specific machine.

186
00:07:25,080 --> 00:07:28,320
Standard times provide a consistent basis for planning and costing,

187
00:07:28,320 --> 00:07:30,000
even when actual conditions vary.

188
00:07:30,000 --> 00:07:31,600
Those assumptions aren't careless.

189
00:07:31,600 --> 00:07:34,520
They make enterprise planning manageable across a complex business.

190
00:07:34,520 --> 00:07:36,560
The trouble starts when people ask those assumptions

191
00:07:36,560 --> 00:07:39,040
to answer a more detailed question than they were built for.

192
00:07:39,040 --> 00:07:41,320
A work order can be completely clean in ERP.

193
00:07:41,320 --> 00:07:43,520
It can contain the right material list,

194
00:07:43,520 --> 00:07:47,360
approved rooting, target quantity, and requested date.

195
00:07:47,360 --> 00:07:49,280
That still doesn't create an empty machine.

196
00:07:49,280 --> 00:07:51,200
ERP tells you that work should happen.

197
00:07:51,200 --> 00:07:54,480
It creates the supply signal, reserves or plans material,

198
00:07:54,480 --> 00:07:56,640
tracks the order and records the result.

199
00:07:56,640 --> 00:08:00,400
It gives the factory a business framework that production planning depends on.

200
00:08:00,400 --> 00:08:05,200
But on a real shop floor, work competes for time, people, equipment, tools, and attention.

201
00:08:05,200 --> 00:08:07,040
Those conditions shift through the day,

202
00:08:07,040 --> 00:08:09,360
and the gap between a controlled business plan

203
00:08:09,360 --> 00:08:11,320
and physical capacity starts to grow.

204
00:08:11,320 --> 00:08:13,120
That's where we need to look next.

205
00:08:13,120 --> 00:08:16,000
Why ERP planning often looks more certain than it is?

206
00:08:16,000 --> 00:08:19,680
A lot of ERP plans look firm because they contain dates, start date,

207
00:08:19,680 --> 00:08:21,360
finish date, due date.

208
00:08:21,360 --> 00:08:24,240
The problem is that a date inside a plan can look like a commitment

209
00:08:24,240 --> 00:08:28,400
even when it came from assumptions that never tested the actual factory conditions.

210
00:08:28,400 --> 00:08:30,640
Capacity is often the first assumption that breaks.

211
00:08:30,640 --> 00:08:34,640
In many ERP planning flows, the system checks whether work belongs in a period,

212
00:08:34,640 --> 00:08:36,080
maybe a day or a week,

213
00:08:36,080 --> 00:08:40,080
without fully asking whether all that work can fit on the same constrained resource.

214
00:08:40,080 --> 00:08:42,080
This is called infinite capacity planning.

215
00:08:42,080 --> 00:08:44,960
It doesn't mean ERP believes machines can run forever.

216
00:08:44,960 --> 00:08:47,360
It means the planning run can place demand against a work center

217
00:08:47,360 --> 00:08:51,040
without forcing every order to compete for a real place in a detailed sequence.

218
00:08:51,040 --> 00:08:52,560
That approach serves a purpose.

219
00:08:52,560 --> 00:08:54,320
It gives planners a broad demand signal.

220
00:08:54,320 --> 00:08:57,120
It can show that a work center carries too much load next month

221
00:08:57,120 --> 00:09:00,240
or that a supply plan needs attention before a due date arrives.

222
00:09:00,240 --> 00:09:04,400
But broad capacity isn't the same as a workable production schedule, not even close.

223
00:09:04,400 --> 00:09:07,040
Consider one work center with a normal daily capacity.

224
00:09:07,040 --> 00:09:09,200
ERP may place several orders into the same day

225
00:09:09,200 --> 00:09:12,880
because each order carries a planned duration and the routing points to that work center.

226
00:09:12,880 --> 00:09:14,320
On paper every order has a date.

227
00:09:14,320 --> 00:09:18,000
On the floor somebody still has to decide the order in which those jobs run,

228
00:09:18,000 --> 00:09:19,440
where the setup time goes,

229
00:09:19,440 --> 00:09:21,680
and which job waits when the day runs out.

230
00:09:21,680 --> 00:09:23,200
The date didn't solve those decisions,

231
00:09:23,200 --> 00:09:25,920
it postponed them to whoever releases the work.

232
00:09:25,920 --> 00:09:28,240
Fixed lead times create a similar effect.

233
00:09:28,240 --> 00:09:32,480
An ERP route may assume an operation takes two days from release to completion.

234
00:09:32,480 --> 00:09:35,680
That duration includes processing time, waiting time inspection,

235
00:09:35,680 --> 00:09:37,840
transport and a buffer for normal variation.

236
00:09:37,840 --> 00:09:40,000
For enterprise planning, that's sensible.

237
00:09:40,000 --> 00:09:43,280
Yet the same two day lead time can hide very different conditions.

238
00:09:43,280 --> 00:09:44,800
One job may run in 30 minutes,

239
00:09:44,800 --> 00:09:47,280
but wait a day and a half because the machine has a queue.

240
00:09:47,280 --> 00:09:49,280
Another may require a long change over

241
00:09:49,280 --> 00:09:52,400
because the previous product uses different material or tooling.

242
00:09:52,400 --> 00:09:55,280
A third may stop because maintenance takes the resource offline

243
00:09:55,280 --> 00:09:58,880
or because the same operator needs to support another machine simultaneously.

244
00:09:58,880 --> 00:10:00,400
The routing still shows two days.

245
00:10:00,400 --> 00:10:03,280
The factory sees a series of real physical constraints.

246
00:10:03,280 --> 00:10:05,520
Shared resources make that gap even wider.

247
00:10:05,520 --> 00:10:09,520
A machine might have open time, but the fixture needed for the job sits on another line.

248
00:10:09,520 --> 00:10:10,880
A test cell may have room,

249
00:10:10,880 --> 00:10:13,280
but the qualified person only works a certain shift.

250
00:10:13,280 --> 00:10:16,400
A process can look available when viewed one resource at a time,

251
00:10:16,400 --> 00:10:19,520
while the combined set of resources cannot support the job at all.

252
00:10:19,520 --> 00:10:21,760
This is why a capacity number needs context,

253
00:10:21,760 --> 00:10:23,360
not just a count of hours.

254
00:10:23,360 --> 00:10:26,000
Bucketed planning hides the same issue in a different way.

255
00:10:26,000 --> 00:10:28,640
If capacity appears by week, 20 hours of work

256
00:10:28,640 --> 00:10:30,960
and 20 hours of capacity can look perfectly balanced.

257
00:10:30,960 --> 00:10:32,640
But where do those hours actually sit?

258
00:10:32,640 --> 00:10:35,600
If all the urgent work needs the same machine on Monday morning,

259
00:10:35,600 --> 00:10:37,520
while most capacity sits later in the week,

260
00:10:37,520 --> 00:10:41,200
the weekly view looks calm and the dispatch problem remains unsolved.

261
00:10:41,200 --> 00:10:42,880
Sequence changes the answer to.

262
00:10:42,880 --> 00:10:46,160
Two orders may both fit within a day, but not in either order.

263
00:10:46,160 --> 00:10:49,120
If one job requires a clean down after a certain product family,

264
00:10:49,120 --> 00:10:51,760
running it first may consume the time needed for the second.

265
00:10:51,760 --> 00:10:53,840
If the second runs first, both may fit.

266
00:10:53,840 --> 00:10:56,640
A weekly bucket can't express that difference with enough detail

267
00:10:56,640 --> 00:10:58,720
to guide the person releasing work.

268
00:10:58,720 --> 00:11:00,640
None of this makes ERP planning useless.

269
00:11:00,640 --> 00:11:03,200
It means we need to give it the job it can actually do well.

270
00:11:03,200 --> 00:11:06,560
MRP works especially well as a demand and supply calculation.

271
00:11:06,560 --> 00:11:09,120
It asks what materials and components the business needs,

272
00:11:09,120 --> 00:11:12,800
when it needs them, and whether planned or existing supply covers that need.

273
00:11:12,800 --> 00:11:14,880
It can surface shortages early and trigger,

274
00:11:14,880 --> 00:11:18,000
purchasing or production signals across a complex bill of materials.

275
00:11:18,000 --> 00:11:19,040
That's its strength.

276
00:11:19,040 --> 00:11:22,160
MRP is not designed to be a minute-by-minute traffic controller

277
00:11:22,160 --> 00:11:24,000
for every job moving through the plant.

278
00:11:24,000 --> 00:11:26,560
Trying to use it that way creates false certainty.

279
00:11:26,560 --> 00:11:28,240
A planarcy's planned completion dates

280
00:11:28,240 --> 00:11:29,760
and assumes the factory can meet them.

281
00:11:29,760 --> 00:11:31,360
Production sees a list of work orders

282
00:11:31,360 --> 00:11:34,000
and still has to sort out the physical order of execution.

283
00:11:34,000 --> 00:11:35,360
Sales sees a due date.

284
00:11:35,360 --> 00:11:38,320
While the bottleneck resource sees a queue it can't clear.

285
00:11:38,320 --> 00:11:40,160
So treat the ERP plan as a target,

286
00:11:40,160 --> 00:11:42,320
a control point, a business agreement.

287
00:11:42,320 --> 00:11:44,560
It tells the organization what demand exists,

288
00:11:44,560 --> 00:11:46,000
what supply needs creating,

289
00:11:46,000 --> 00:11:47,760
and which transactions need governing.

290
00:11:47,760 --> 00:11:50,880
It doesn't, by itself, prove that the schedule can actually run.

291
00:11:50,880 --> 00:11:54,560
The next question is what changes when the planning system treats machines,

292
00:11:54,560 --> 00:11:56,640
people, tools, material timing,

293
00:11:56,640 --> 00:12:00,320
and sequence as active constraints rather than background assumptions.

294
00:12:00,320 --> 00:12:03,600
That's where advance planning and scheduling enters the same problem

295
00:12:03,600 --> 00:12:05,760
and where the real conversation starts.

296
00:12:05,760 --> 00:12:07,680
What APS software actually does.

297
00:12:07,680 --> 00:12:11,440
So here's what APS, advance planning and scheduling actually does.

298
00:12:11,440 --> 00:12:14,400
It takes the same demand and production data you already have

299
00:12:14,400 --> 00:12:17,280
and turns it into a schedule that respects real-world limits.

300
00:12:17,280 --> 00:12:19,280
That sounds more complicated than it really is.

301
00:12:19,280 --> 00:12:21,440
Basically, the schedule has to deal with the constraints

302
00:12:21,440 --> 00:12:23,120
that exist in your plant right now.

303
00:12:23,120 --> 00:12:24,720
A machine has a real calendar,

304
00:12:24,720 --> 00:12:26,400
an operator works a real shift,

305
00:12:26,400 --> 00:12:28,880
a tool can only be in one place at a time.

306
00:12:28,880 --> 00:12:32,000
Material arrives when it arrives, not when you wish it did.

307
00:12:32,000 --> 00:12:34,720
And some products can follow each other with almost no setup,

308
00:12:34,720 --> 00:12:37,600
while others force a long change over, a cleaning step,

309
00:12:37,600 --> 00:12:39,760
or an inspection before the next job starts.

310
00:12:39,760 --> 00:12:43,520
APS brings all those conditions into the planning calculation itself.

311
00:12:43,520 --> 00:12:45,840
Think of the schedule as a sequence of operations

312
00:12:45,840 --> 00:12:47,680
moving through limited time slots.

313
00:12:47,680 --> 00:12:50,160
Each operation needs a place to run, a duration,

314
00:12:50,160 --> 00:12:52,720
and a set of conditions before it can even start.

315
00:12:52,720 --> 00:12:55,120
APS checks those conditions as it assigns work,

316
00:12:55,120 --> 00:12:57,520
instead of just dropping every order onto its planned date

317
00:12:57,520 --> 00:13:00,000
and leaving the collision for somebody else to clean up.

318
00:13:00,000 --> 00:13:02,400
Finite capacity is the core of this approach.

319
00:13:02,400 --> 00:13:04,960
If a machine only has six available hours in a shift,

320
00:13:04,960 --> 00:13:08,320
APS can't honestly load 10 hours of work into that same shift

321
00:13:08,320 --> 00:13:09,360
and call it a schedule.

322
00:13:09,360 --> 00:13:12,480
It has to move work, take an alternate resource if one exists,

323
00:13:12,480 --> 00:13:14,080
push the plan into another period,

324
00:13:14,080 --> 00:13:16,240
or flag a conflict that needs a human decision.

325
00:13:16,240 --> 00:13:18,240
That doesn't mean the system blocks every exception.

326
00:13:18,240 --> 00:13:19,520
Factories run on exceptions.

327
00:13:19,520 --> 00:13:21,360
You might choose to add overtime,

328
00:13:21,360 --> 00:13:23,840
postpone PM, authorize a subcontractor,

329
00:13:23,840 --> 00:13:25,760
or accept that one order will ship late.

330
00:13:25,760 --> 00:13:28,240
The difference is that APS makes the trade off visible.

331
00:13:28,240 --> 00:13:31,280
Instead of quietly stuffing work into capacity that isn't there,

332
00:13:31,280 --> 00:13:33,520
the system shows you the consequence of your choice.

333
00:13:33,520 --> 00:13:35,360
Pull one urgent order forward,

334
00:13:35,360 --> 00:13:36,800
which orders get pushed back.

335
00:13:36,800 --> 00:13:38,080
Select an alternate machine.

336
00:13:38,080 --> 00:13:39,680
Does that change processing time

337
00:13:39,680 --> 00:13:42,240
or require a qualified operator run overtime?

338
00:13:42,240 --> 00:13:43,760
Does the order actually finish earlier

339
00:13:43,760 --> 00:13:45,520
or does material still hold it up?

340
00:13:45,520 --> 00:13:47,360
That turns scheduling into a decision process,

341
00:13:47,360 --> 00:13:48,800
not just a date calculation.

342
00:13:48,800 --> 00:13:51,600
Now, sequence matters just as much as raw capacity.

343
00:13:51,600 --> 00:13:53,200
A planner rarely asks only,

344
00:13:53,200 --> 00:13:54,640
"Do we have eight hours free?"

345
00:13:54,640 --> 00:13:56,560
They ask, "What should run next?"

346
00:13:56,560 --> 00:13:59,840
APS can apply rules that reflect how your plant actually works,

347
00:13:59,840 --> 00:14:02,080
due date may drive one part of the sequence.

348
00:14:02,080 --> 00:14:03,920
Setup family may drive another.

349
00:14:03,920 --> 00:14:06,080
If several orders use the same material,

350
00:14:06,080 --> 00:14:08,560
tool, color, grade, or process setting,

351
00:14:08,560 --> 00:14:11,040
grouping them together reduces changeovers.

352
00:14:11,040 --> 00:14:12,160
In process industries,

353
00:14:12,160 --> 00:14:14,000
planners often use campaign logic

354
00:14:14,000 --> 00:14:15,440
where similar products run together

355
00:14:15,440 --> 00:14:18,400
for a set period before the line switches to another family.

356
00:14:18,400 --> 00:14:19,840
Priority also enters the picture,

357
00:14:19,840 --> 00:14:21,280
but it needs some care.

358
00:14:21,280 --> 00:14:22,640
A priority flag sounds simple

359
00:14:22,640 --> 00:14:25,280
until every sales order carries the highest priority.

360
00:14:25,280 --> 00:14:28,320
APS can't fix a priority policy that nobody controls.

361
00:14:28,320 --> 00:14:30,560
It can apply the rules you give it and show the result,

362
00:14:30,560 --> 00:14:32,480
but the business still has to decide

363
00:14:32,480 --> 00:14:34,080
whether a late customer order

364
00:14:34,080 --> 00:14:36,720
outweighs the cost of breaking a production campaign

365
00:14:36,720 --> 00:14:38,160
or delaying other work.

366
00:14:38,160 --> 00:14:41,520
Material timing changes the sequence, too.

367
00:14:41,520 --> 00:14:42,880
An order might look urgent,

368
00:14:42,880 --> 00:14:45,120
but if a required component arrives tomorrow afternoon,

369
00:14:45,120 --> 00:14:47,680
scheduling it first this morning accomplishes almost nothing,

370
00:14:47,680 --> 00:14:49,840
a constraint-based schedule can place work

371
00:14:49,840 --> 00:14:51,680
where it has a chance of actually starting,

372
00:14:51,680 --> 00:14:53,360
rather than filling the front of the queue

373
00:14:53,360 --> 00:14:55,680
with jobs that just sit there waiting for parts.

374
00:14:55,680 --> 00:14:59,120
Then the factory changes, because factories always do.

375
00:14:59,120 --> 00:15:00,320
A machine goes down,

376
00:15:00,320 --> 00:15:01,680
a supplier delivery slips,

377
00:15:01,680 --> 00:15:03,040
an operator calls in sick,

378
00:15:03,040 --> 00:15:04,720
a customer changes a requirement,

379
00:15:04,720 --> 00:15:06,320
a job finishes early,

380
00:15:06,320 --> 00:15:07,440
or takes twice as long.

381
00:15:07,440 --> 00:15:09,920
APS can recalculate the schedule

382
00:15:09,920 --> 00:15:11,360
when those conditions change

383
00:15:11,360 --> 00:15:12,560
and test a new arrangement

384
00:15:12,560 --> 00:15:14,800
against the rules and constraints you've already defined

385
00:15:14,800 --> 00:15:16,400
that doesn't mean every event should trigger

386
00:15:16,400 --> 00:15:17,840
an uncontrolled reshuffle,

387
00:15:17,840 --> 00:15:19,440
a schedule that changes every few minutes

388
00:15:19,440 --> 00:15:20,800
becomes impossible to run.

389
00:15:20,800 --> 00:15:24,080
But APS gives plans the ability to see the effect of a change quickly,

390
00:15:24,080 --> 00:15:25,120
compare options,

391
00:15:25,120 --> 00:15:28,160
and decide when a revised plan should become the new dispatch order.

392
00:15:28,160 --> 00:15:30,800
The output isn't just a prettier list of work orders,

393
00:15:30,800 --> 00:15:33,440
a useful APS result can show a feasible sequence

394
00:15:33,440 --> 00:15:34,560
for each resource.

395
00:15:34,560 --> 00:15:36,800
The plans start and finish of each operation,

396
00:15:36,800 --> 00:15:38,240
any unresolved conflicts,

397
00:15:38,240 --> 00:15:40,560
and the impact on due dates when priorities shift.

398
00:15:40,560 --> 00:15:43,360
It can also produce a ranked dispatch list,

399
00:15:43,360 --> 00:15:45,440
so a supervisor knows which jobs should run next

400
00:15:45,440 --> 00:15:46,560
under the current plan.

401
00:15:46,560 --> 00:15:48,960
More importantly, it can show you alternatives.

402
00:15:48,960 --> 00:15:50,480
Maybe the urgent order finishes on time

403
00:15:50,480 --> 00:15:52,160
if it runs on an alternate machine.

404
00:15:52,160 --> 00:15:53,280
Maybe it finishes on time,

405
00:15:53,280 --> 00:15:55,040
only if production accepts overtime.

406
00:15:55,040 --> 00:15:56,560
Maybe it can't finish on time

407
00:15:56,560 --> 00:15:58,000
under any approved condition,

408
00:15:58,000 --> 00:15:59,520
and the plan needs to tell sales

409
00:15:59,520 --> 00:16:01,840
before another unrealistic date enters the system.

410
00:16:01,840 --> 00:16:03,280
That's what APS really adds.

411
00:16:03,280 --> 00:16:05,520
It doesn't remove uncertainty from the factory.

412
00:16:05,520 --> 00:16:07,520
It gives the organization a structured way

413
00:16:07,520 --> 00:16:09,360
to test uncertainty against known limits

414
00:16:09,360 --> 00:16:11,200
before people act on a plan.

415
00:16:11,200 --> 00:16:13,200
Keep the word "feasible" in mind, though,

416
00:16:13,200 --> 00:16:14,400
because a feasible schedule

417
00:16:14,400 --> 00:16:17,120
and the best possible schedule aren't always the same thing.

418
00:16:17,120 --> 00:16:18,880
Feasible does not mean perfect.

419
00:16:18,880 --> 00:16:20,080
A schedule can be feasible

420
00:16:20,080 --> 00:16:21,840
and still be a terrible business choice.

421
00:16:21,840 --> 00:16:24,640
Feasibility just means the schedule respects the constraints,

422
00:16:24,640 --> 00:16:25,440
it knows about.

423
00:16:25,440 --> 00:16:27,840
It doesn't put the same machine on two jobs at once.

424
00:16:27,840 --> 00:16:30,880
It doesn't schedule a job before it's required material arrives.

425
00:16:30,880 --> 00:16:33,120
It accounts for the resource rules, timing rules,

426
00:16:33,120 --> 00:16:34,880
and root logic in the planning model,

427
00:16:34,880 --> 00:16:36,160
and it exposes conflicts

428
00:16:36,160 --> 00:16:39,120
when the plant can't meet all demand under those conditions.

429
00:16:39,120 --> 00:16:40,880
That gives you an honest starting point.

430
00:16:40,880 --> 00:16:43,520
Imagine a schedule where every order fits inside

431
00:16:43,520 --> 00:16:44,800
available capacity,

432
00:16:44,800 --> 00:16:46,880
every operation follows its approved root,

433
00:16:46,880 --> 00:16:49,200
and no qualified operator is double booked.

434
00:16:49,200 --> 00:16:50,960
The system can call that schedule feasible.

435
00:16:50,960 --> 00:16:53,920
But it may still create too much work in process,

436
00:16:53,920 --> 00:16:56,640
too much partly finished stock sitting between operations.

437
00:16:56,640 --> 00:16:58,320
It may cause frequent setups,

438
00:16:58,320 --> 00:16:59,920
or it may protect every due date,

439
00:16:59,920 --> 00:17:02,560
except the one customer order the business cares about most.

440
00:17:02,560 --> 00:17:04,800
A feasible schedule answers,

441
00:17:04,800 --> 00:17:05,680
can this run?

442
00:17:05,680 --> 00:17:07,680
Optimization asks the different question.

443
00:17:07,680 --> 00:17:09,200
Out of the schedules that can run,

444
00:17:09,200 --> 00:17:11,200
which one produces the result we want,

445
00:17:11,200 --> 00:17:12,480
that sounds like a small difference.

446
00:17:12,480 --> 00:17:14,640
In practice, it changes the whole conversation.

447
00:17:14,640 --> 00:17:18,240
An optimization engine searches through possible schedules using stated goals.

448
00:17:18,240 --> 00:17:19,760
You might want better delivery performance,

449
00:17:19,760 --> 00:17:21,760
so the system gives more weight to orders

450
00:17:21,760 --> 00:17:23,840
at risk of missing their committed date.

451
00:17:23,840 --> 00:17:26,080
You might need higher throughput through a bottleneck,

452
00:17:26,080 --> 00:17:29,120
so the plant protects the resource that limits overall output.

453
00:17:29,120 --> 00:17:30,560
Or you might want fewer setups

454
00:17:30,560 --> 00:17:32,400
because long change over is each time,

455
00:17:32,400 --> 00:17:33,520
creates scrap risk,

456
00:17:33,520 --> 00:17:36,240
and leave people constantly preparing rather than producing.

457
00:17:36,240 --> 00:17:38,720
Each goal pushes the schedule in a different direction.

458
00:17:38,720 --> 00:17:40,720
Lower work in process is another common goal.

459
00:17:40,720 --> 00:17:42,240
If jobs enter production too early,

460
00:17:42,240 --> 00:17:43,600
they sit between steps for days.

461
00:17:43,600 --> 00:17:45,920
That ties up material, hides problems,

462
00:17:45,920 --> 00:17:47,360
and makes the plant look busy,

463
00:17:47,360 --> 00:17:49,680
even when finished goods don't leave any faster.

464
00:17:49,680 --> 00:17:51,360
A schedule can control release timings,

465
00:17:51,360 --> 00:17:54,400
so work arrives closer to when the next operation can actually start.

466
00:17:54,400 --> 00:17:57,040
But there isn't one universal best schedule.

467
00:17:57,040 --> 00:17:58,560
Think about an urgent customer order

468
00:17:58,560 --> 00:18:01,840
that needs a machine currently set up for a family of similar products.

469
00:18:01,840 --> 00:18:03,120
Keeping that family together,

470
00:18:03,120 --> 00:18:06,080
may cut setup time and preserve output for the shift.

471
00:18:06,080 --> 00:18:08,960
Pulling the urgent order forward may require a change over,

472
00:18:08,960 --> 00:18:11,360
then another change over to get back to the original family.

473
00:18:11,360 --> 00:18:12,720
One choice improves flow.

474
00:18:12,720 --> 00:18:14,160
The other protects a customer promise.

475
00:18:14,160 --> 00:18:15,760
Neither choice is automatically wrong.

476
00:18:15,760 --> 00:18:17,360
The system can calculate the impact.

477
00:18:17,360 --> 00:18:18,960
It can show which orders move,

478
00:18:18,960 --> 00:18:20,960
how much setup time the change needs,

479
00:18:20,960 --> 00:18:22,800
whether the urgent order meets its date,

480
00:18:22,800 --> 00:18:24,480
and where the queue shifts afterward.

481
00:18:24,480 --> 00:18:27,120
But it can't decide what the business should care about

482
00:18:27,120 --> 00:18:29,360
unless the business has already defined that policy.

483
00:18:29,360 --> 00:18:31,520
This is where many planning projects get awkward.

484
00:18:31,520 --> 00:18:33,200
People ask the system to optimize,

485
00:18:33,200 --> 00:18:34,800
then discover that sales, production,

486
00:18:34,800 --> 00:18:37,600
and supply chain never agreed on the objective.

487
00:18:37,600 --> 00:18:40,240
Sales once every committed date protected.

488
00:18:40,240 --> 00:18:42,800
Production once stable campaigns and fewer changes.

489
00:18:42,800 --> 00:18:44,720
Supply chain once lower inventory.

490
00:18:44,720 --> 00:18:48,720
Finance may care about margin, expedited freight, or stock exposure.

491
00:18:48,720 --> 00:18:50,240
Each position is reasonable,

492
00:18:50,240 --> 00:18:52,560
but they can pull the schedule in opposite directions.

493
00:18:52,560 --> 00:18:54,880
No amount of better math can settle that argument.

494
00:18:54,880 --> 00:18:55,520
You need rules.

495
00:18:55,520 --> 00:18:58,480
Those rules might state that certain customers receive priority

496
00:18:58,480 --> 00:19:01,360
that a safety or quality limit overrides delivery pressure

497
00:19:01,360 --> 00:19:03,120
or that the schedule should stay stable

498
00:19:03,120 --> 00:19:06,720
inside a defined time window unless a planner approves the change.

499
00:19:06,720 --> 00:19:09,280
They might also define how much lateness the business accepts

500
00:19:09,280 --> 00:19:11,520
before it breaks a campaign or adds overtime.

501
00:19:11,520 --> 00:19:14,640
That's human policy translated into scheduling logic.

502
00:19:14,640 --> 00:19:17,040
Without it, optimization becomes a fast way

503
00:19:17,040 --> 00:19:19,280
to produce a plan that nobody trusts.

504
00:19:19,280 --> 00:19:21,440
The schedule may look mathematically tidy

505
00:19:21,440 --> 00:19:23,040
while ignoring the commercial decision

506
00:19:23,040 --> 00:19:24,880
a planner would have made in 10 seconds.

507
00:19:24,880 --> 00:19:27,360
Or it may keep changing because the model treats

508
00:19:27,360 --> 00:19:30,160
every small improvement as more important than a stable plan

509
00:19:30,160 --> 00:19:31,920
people can actually execute.

510
00:19:31,920 --> 00:19:35,040
A good APS process doesn't hand authority to an algorithm.

511
00:19:35,040 --> 00:19:38,000
It gives planners a clearer view of the choices in front of them

512
00:19:38,000 --> 00:19:40,080
along with the consequences of each choice.

513
00:19:40,080 --> 00:19:41,600
That changes the planner's work.

514
00:19:41,600 --> 00:19:44,240
Instead of spending hours trying to build a possible sequence

515
00:19:44,240 --> 00:19:45,440
from scattered facts,

516
00:19:45,440 --> 00:19:47,280
the planner can challenge the result.

517
00:19:47,280 --> 00:19:48,560
Why did this order move?

518
00:19:48,560 --> 00:19:50,160
Which rule caused that conflict?

519
00:19:50,160 --> 00:19:52,240
What happens if we protect this customer data

520
00:19:52,240 --> 00:19:54,080
and accept a later finish elsewhere?

521
00:19:54,080 --> 00:19:55,600
Those are real planning decisions.

522
00:19:55,600 --> 00:19:56,400
They need judgment.

523
00:19:56,400 --> 00:19:59,280
So let's leave the scheduling theory for a moment

524
00:19:59,280 --> 00:20:01,360
and put it under pressure on the factory floor.

525
00:20:01,360 --> 00:20:03,840
Picture a Monday morning where capacity disappears,

526
00:20:03,840 --> 00:20:05,280
a customer escalates,

527
00:20:05,280 --> 00:20:06,960
and the plan has to survive

528
00:20:06,960 --> 00:20:08,880
more than a clean set of assumptions.

529
00:20:08,880 --> 00:20:10,800
The Monday morning disruption test.

530
00:20:10,800 --> 00:20:12,960
Picture a plant at the start of a Monday shift.

531
00:20:12,960 --> 00:20:15,200
A machining center handles a critical operation

532
00:20:15,200 --> 00:20:16,560
for several active orders

533
00:20:16,560 --> 00:20:18,000
and right after the shift kicks off

534
00:20:18,000 --> 00:20:19,440
maintenance confirms a fault.

535
00:20:19,440 --> 00:20:21,280
That machine is down for the rest of the day.

536
00:20:21,280 --> 00:20:22,480
Almost at the same time,

537
00:20:22,480 --> 00:20:23,840
a customer escalates an order

538
00:20:23,840 --> 00:20:26,080
that already carries a tight delivery date.

539
00:20:26,080 --> 00:20:28,480
Sales wants to know whether production can still bring it forward.

540
00:20:28,480 --> 00:20:30,720
There's an alternate machine that can run it.

541
00:20:30,720 --> 00:20:31,600
But it's not empty,

542
00:20:31,600 --> 00:20:34,240
isolated is already running a different product family

543
00:20:34,240 --> 00:20:36,080
and only one operator on that shift

544
00:20:36,080 --> 00:20:37,920
can handle the urgent order there.

545
00:20:37,920 --> 00:20:40,320
This is where planning stops being a date exercise.

546
00:20:40,320 --> 00:20:42,480
In ERP, the planner can see the order status.

547
00:20:42,480 --> 00:20:43,680
It's planned start and finish,

548
00:20:43,680 --> 00:20:45,680
the due date and the material position.

549
00:20:45,680 --> 00:20:48,000
They can check whether components are still on hand,

550
00:20:48,000 --> 00:20:50,960
whether purchase orders still support later operations,

551
00:20:50,960 --> 00:20:52,480
and whether a shortage might appear

552
00:20:52,480 --> 00:20:53,920
if the sequence changes.

553
00:20:53,920 --> 00:20:56,400
That view gives the business real control.

554
00:20:56,400 --> 00:20:59,040
The planner knows which order belongs to which customer,

555
00:20:59,040 --> 00:21:00,480
which work orders are open,

556
00:21:00,480 --> 00:21:03,040
and where supply signals may need attention.

557
00:21:03,040 --> 00:21:04,320
Sales, purchasing, warehouse,

558
00:21:04,320 --> 00:21:06,880
and finance all work from the same transaction record.

559
00:21:06,880 --> 00:21:08,640
But here's what that doesn't answer.

560
00:21:08,640 --> 00:21:10,880
What can the plant actually do this afternoon?

561
00:21:10,880 --> 00:21:12,800
The ERP dates may still show a normal plan

562
00:21:12,800 --> 00:21:15,120
because the disruption hasn't hit the order records yet.

563
00:21:15,120 --> 00:21:16,720
Even after someone changes a date,

564
00:21:16,720 --> 00:21:18,640
the system won't show the full chain of work

565
00:21:18,640 --> 00:21:20,080
displaced by the machine failure.

566
00:21:20,080 --> 00:21:22,000
A planned date tells you the target,

567
00:21:22,000 --> 00:21:24,160
but it doesn't automatically test every consequence

568
00:21:24,160 --> 00:21:26,400
of losing a real resource for a real shift.

569
00:21:26,400 --> 00:21:27,680
APS approaches the problem

570
00:21:27,680 --> 00:21:29,840
from the affected operations outward.

571
00:21:29,840 --> 00:21:31,760
First, it identifies the operations scheduled

572
00:21:31,760 --> 00:21:33,040
on the failed machine.

573
00:21:33,040 --> 00:21:36,640
Then it checks whether each operation has an approved alternate resource,

574
00:21:36,640 --> 00:21:38,880
whether that resource has enough free time,

575
00:21:38,880 --> 00:21:40,480
and whether the job needs a tool,

576
00:21:40,480 --> 00:21:42,960
setup, material, or qualified operator

577
00:21:42,960 --> 00:21:44,560
that changes the answer.

578
00:21:44,560 --> 00:21:46,720
The customer escalation adds more pressure.

579
00:21:46,720 --> 00:21:49,120
The planner isn't just recovering lost capacity.

580
00:21:49,120 --> 00:21:51,280
They're also testing whether one order can move ahead

581
00:21:51,280 --> 00:21:52,880
of the work already in the queue.

582
00:21:52,880 --> 00:21:55,680
Suppose the urgent order can run on the alternate machine.

583
00:21:55,680 --> 00:21:57,840
APS can test the revised sequence.

584
00:21:57,840 --> 00:22:00,320
It can show the setup needed before that order starts,

585
00:22:00,320 --> 00:22:01,840
the orders that move later,

586
00:22:01,840 --> 00:22:03,680
and the effect on their completion dates.

587
00:22:03,680 --> 00:22:05,600
If the operator needed for the urgent order

588
00:22:05,600 --> 00:22:07,200
also supports another process,

589
00:22:07,200 --> 00:22:09,120
the model exposes that conflict,

590
00:22:09,120 --> 00:22:10,480
instead of leaving the supervisor

591
00:22:10,480 --> 00:22:12,480
to discover it halfway through the shift.

592
00:22:12,480 --> 00:22:14,160
That changes the conversation.

593
00:22:14,160 --> 00:22:15,120
Instead of asking,

594
00:22:15,120 --> 00:22:17,120
can we fit this urgent order in?

595
00:22:17,120 --> 00:22:18,240
The planner can now ask,

596
00:22:18,240 --> 00:22:20,400
if we fit it in, who loses capacity?

597
00:22:20,400 --> 00:22:21,840
Those are very different questions.

598
00:22:21,840 --> 00:22:23,840
The first often leads to a quick promise.

599
00:22:23,840 --> 00:22:26,720
The second forces the business to see the cost of that promise,

600
00:22:26,720 --> 00:22:28,080
which customer dates move,

601
00:22:28,080 --> 00:22:29,520
because the machine went down,

602
00:22:29,520 --> 00:22:31,200
which jobs can move to another resource

603
00:22:31,200 --> 00:22:33,600
without creating a quality or labor issue.

604
00:22:33,600 --> 00:22:36,080
Does the alternate machine become the new bottleneck

605
00:22:36,080 --> 00:22:38,640
once it absorbs work from the failed machine?

606
00:22:38,640 --> 00:22:40,400
And does protecting the urgent order

607
00:22:40,400 --> 00:22:42,160
push a problem into a later operation

608
00:22:42,160 --> 00:22:43,840
where it becomes harder to recover?

609
00:22:43,840 --> 00:22:46,160
A planning system can't remove the lost hours.

610
00:22:46,160 --> 00:22:47,760
Nobody expects it to.

611
00:22:47,760 --> 00:22:49,600
What it can do is turn a vague emergency

612
00:22:49,600 --> 00:22:51,120
into a set of visible choices.

613
00:22:51,120 --> 00:22:52,800
Run over time, move selected orders,

614
00:22:52,800 --> 00:22:54,000
protect one customer date

615
00:22:54,000 --> 00:22:55,440
and accept the delay elsewhere,

616
00:22:55,440 --> 00:22:56,720
or leave the sequence stable

617
00:22:56,720 --> 00:22:58,800
and tell sales the earliest on a state.

618
00:22:58,800 --> 00:23:00,160
Each option has a consequence

619
00:23:00,160 --> 00:23:02,000
and the planner needs to see that consequence

620
00:23:02,000 --> 00:23:03,440
before the business commits.

621
00:23:03,440 --> 00:23:04,400
Without that support,

622
00:23:04,400 --> 00:23:06,400
the familiar rescue process kicks in.

623
00:23:06,400 --> 00:23:08,160
Someone exports orders from ERP,

624
00:23:08,160 --> 00:23:09,920
someone else checks machine availability

625
00:23:09,920 --> 00:23:11,520
from memory or local board,

626
00:23:11,520 --> 00:23:13,520
a supervisor calls the shift lead,

627
00:23:13,520 --> 00:23:14,960
and the planner builds a spreadsheet

628
00:23:14,960 --> 00:23:17,280
that answers today's question while purchasing,

629
00:23:17,280 --> 00:23:18,560
sales and production,

630
00:23:18,560 --> 00:23:20,880
each wait for a different version of the new plan?

631
00:23:20,880 --> 00:23:24,720
Spreadsheets aren't the problem by themselves.

632
00:23:24,720 --> 00:23:27,120
Under pressure, they can be fast and practical.

633
00:23:27,120 --> 00:23:28,160
The rule problem is,

634
00:23:28,160 --> 00:23:29,760
when the spreadsheet becomes the only place

635
00:23:29,760 --> 00:23:31,520
where current factory logic exists,

636
00:23:31,520 --> 00:23:33,520
with no clear link back to the orders,

637
00:23:33,520 --> 00:23:34,320
constraints,

638
00:23:34,320 --> 00:23:36,160
and decisions that everyone else needs to trust.

639
00:23:36,160 --> 00:23:38,000
Controlled re-planning means

640
00:23:38,000 --> 00:23:39,280
the planner can test scenarios

641
00:23:39,280 --> 00:23:41,440
without releasing every draft schedule to the floor,

642
00:23:41,440 --> 00:23:44,160
it means the business agrees on who can change priorities

643
00:23:44,160 --> 00:23:45,760
when a new schedule becomes active

644
00:23:45,760 --> 00:23:48,720
and how resulting dates flow back to the people who need them.

645
00:23:48,720 --> 00:23:50,000
That control matters most

646
00:23:50,000 --> 00:23:51,600
when conditions change quickly,

647
00:23:51,600 --> 00:23:52,640
a machine failure,

648
00:23:52,640 --> 00:23:53,920
an urgent customer request,

649
00:23:53,920 --> 00:23:55,520
and a constrained alternate resource

650
00:23:55,520 --> 00:23:58,080
can all arrive before the first production meeting ends.

651
00:23:58,080 --> 00:24:00,960
ERP and APS each contribute a different part of the answer.

652
00:24:00,960 --> 00:24:02,800
One holds the governed business facts,

653
00:24:02,800 --> 00:24:03,520
the other tests,

654
00:24:03,520 --> 00:24:05,360
the physical consequence of changing the plan,

655
00:24:05,360 --> 00:24:06,720
but for either system to answer,

656
00:24:06,720 --> 00:24:08,160
honestly it needs the right inputs.

657
00:24:08,160 --> 00:24:11,280
So next, let's look at the data each system depends on.

658
00:24:11,280 --> 00:24:12,240
ERP data,

659
00:24:12,240 --> 00:24:13,840
the commercial and transaction layer,

660
00:24:13,840 --> 00:24:15,440
so what kind of data does ERP own?

661
00:24:15,440 --> 00:24:18,240
Start with the business facts around demand,

662
00:24:18,240 --> 00:24:19,840
forecasts, customer orders,

663
00:24:19,840 --> 00:24:21,200
requested delivery dates,

664
00:24:21,200 --> 00:24:22,080
confirmed dates,

665
00:24:22,080 --> 00:24:23,600
and the contract terms that affect

666
00:24:23,600 --> 00:24:25,280
how an order gets handled.

667
00:24:25,280 --> 00:24:27,520
A customer order isn't just a quantity request,

668
00:24:27,520 --> 00:24:29,200
it includes shipped to location,

669
00:24:29,200 --> 00:24:30,880
pricing, delivery rules,

670
00:24:30,880 --> 00:24:32,960
a priority you've agreed with the customer,

671
00:24:32,960 --> 00:24:35,440
and sometimes approval or quality requirements

672
00:24:35,440 --> 00:24:37,520
that follow the product through the business.

673
00:24:37,520 --> 00:24:38,880
Those facts need control,

674
00:24:38,880 --> 00:24:40,640
sales, planning, purchasing,

675
00:24:40,640 --> 00:24:43,840
warehouse, and finance can't each hold their own local version.

676
00:24:43,840 --> 00:24:46,320
ERP gives the order one governed identity,

677
00:24:46,320 --> 00:24:47,760
demand then drives supply.

678
00:24:47,760 --> 00:24:49,920
When the business needs to build or buy something,

679
00:24:49,920 --> 00:24:52,000
ERP connects that demand to inventory,

680
00:24:52,000 --> 00:24:54,880
open purchase orders, production orders, and planned orders.

681
00:24:54,880 --> 00:24:56,720
A planned order is a planning proposal,

682
00:24:56,720 --> 00:24:58,960
it signals that supply needs to exist.

683
00:24:58,960 --> 00:25:00,320
Depending on the process,

684
00:25:00,320 --> 00:25:02,800
it may later become a production order for the plant,

685
00:25:02,800 --> 00:25:05,040
or a purchasing action for an external supplier.

686
00:25:05,040 --> 00:25:09,280
That chain matters because a manufacturer needs to know more than whether a part

687
00:25:09,280 --> 00:25:10,160
should exist.

688
00:25:10,160 --> 00:25:13,040
You also need to know whether the part belongs to stock,

689
00:25:13,040 --> 00:25:14,640
whether it's already committed elsewhere,

690
00:25:14,640 --> 00:25:16,320
whether a supplier has promised it,

691
00:25:16,320 --> 00:25:18,800
and whether production has authority to consume it.

692
00:25:18,800 --> 00:25:20,000
Ownership matters too.

693
00:25:20,000 --> 00:25:22,080
A quantity may sit physically in a warehouse,

694
00:25:22,080 --> 00:25:23,920
but it could belong to a specific customer,

695
00:25:23,920 --> 00:25:25,520
sit in quarantine, wait for inspection,

696
00:25:25,520 --> 00:25:27,040
or remain blocked until approved.

697
00:25:27,040 --> 00:25:30,720
ERP captures those commercial and control states.

698
00:25:30,720 --> 00:25:33,200
Without them, planning confuses physical presence

699
00:25:33,200 --> 00:25:34,480
with usable supply,

700
00:25:34,480 --> 00:25:35,920
and those are not the same thing.

701
00:25:35,920 --> 00:25:37,600
Then there's the product definition.

702
00:25:37,600 --> 00:25:40,000
ERP usually holds the bill of materials,

703
00:25:40,000 --> 00:25:40,880
the bomb.

704
00:25:40,880 --> 00:25:43,120
The bomb tells the business which components,

705
00:25:43,120 --> 00:25:44,880
raw materials, packaging,

706
00:25:44,880 --> 00:25:46,800
or subassemblies belong to a product.

707
00:25:46,800 --> 00:25:48,080
It also holds approved routes

708
00:25:48,080 --> 00:25:51,360
which define the production path that the business recognizes for that item.

709
00:25:51,360 --> 00:25:53,440
Standard times often sit with the route.

710
00:25:53,440 --> 00:25:55,840
Work centers identify the broad production area

711
00:25:55,840 --> 00:25:58,160
or resource group where an operation belongs.

712
00:25:58,160 --> 00:26:00,880
This information supports supply planning, order release,

713
00:26:00,880 --> 00:26:02,080
cost calculation,

714
00:26:02,080 --> 00:26:04,080
and the basic structure of production control.

715
00:26:04,080 --> 00:26:06,000
It also creates a shared language.

716
00:26:06,000 --> 00:26:08,080
When engineering changes a component,

717
00:26:08,080 --> 00:26:10,080
when purchasing sources and alternative,

718
00:26:10,080 --> 00:26:12,240
or when production meets an approved route,

719
00:26:12,240 --> 00:26:15,280
ERP provides a controlled place to manage that change.

720
00:26:15,280 --> 00:26:16,720
You don't want the bill of materials

721
00:26:16,720 --> 00:26:19,120
for a customer order living in a planner's workbook,

722
00:26:19,120 --> 00:26:21,600
especially when traceability, cost,

723
00:26:21,600 --> 00:26:24,560
and quality depend on knowing exactly what the business approved.

724
00:26:24,560 --> 00:26:27,760
Cost is another piece people sometimes forget

725
00:26:27,760 --> 00:26:29,040
during a scheduling discussion.

726
00:26:29,040 --> 00:26:32,160
ERP connects materials, labor assumptions,

727
00:26:32,160 --> 00:26:33,200
work orders, receipts,

728
00:26:33,200 --> 00:26:35,040
and inventory movements to financial records

729
00:26:35,040 --> 00:26:36,400
that supports product costing,

730
00:26:36,400 --> 00:26:37,360
variance analysis,

731
00:26:37,360 --> 00:26:39,360
valuation, invoicing, and financial clothes.

732
00:26:39,360 --> 00:26:41,920
A schedule may decide when work should run,

733
00:26:41,920 --> 00:26:44,400
but the business still needs to know what it consumed,

734
00:26:44,400 --> 00:26:45,120
what it produced,

735
00:26:45,120 --> 00:26:47,120
and what that activity means in financial terms.

736
00:26:47,120 --> 00:26:49,280
Traceability follows the same pattern.

737
00:26:49,280 --> 00:26:50,480
For many manufacturers,

738
00:26:50,480 --> 00:26:53,120
lot and batch information needs to follow material

739
00:26:53,120 --> 00:26:55,600
from receipt through production and shipment.

740
00:26:55,600 --> 00:26:58,000
A business may need to identify which incoming batch

741
00:26:58,000 --> 00:26:59,120
enter the finished product,

742
00:26:59,120 --> 00:27:00,480
which work order process it,

743
00:27:00,480 --> 00:27:01,760
who approved the release,

744
00:27:01,760 --> 00:27:03,920
or which customer received the resulting lot.

745
00:27:03,920 --> 00:27:06,720
ERP often owns much of that formal transaction chain,

746
00:27:06,720 --> 00:27:09,200
sometimes alongside an MES or quality system.

747
00:27:09,200 --> 00:27:10,640
This isn't glamorous work,

748
00:27:10,640 --> 00:27:12,240
it's the part that prevents the factory

749
00:27:12,240 --> 00:27:14,560
from becoming a collection of disconnected actions

750
00:27:14,560 --> 00:27:16,080
with no accountable history.

751
00:27:16,080 --> 00:27:18,880
So when I call ERP, the commercial and transaction layer,

752
00:27:18,880 --> 00:27:21,360
I don't mean it only records things after they happen.

753
00:27:21,360 --> 00:27:24,400
ERP can plan supply, create orders, set dates,

754
00:27:24,400 --> 00:27:25,360
manage approvals,

755
00:27:25,360 --> 00:27:27,760
and connect the demand signal to the money and material

756
00:27:27,760 --> 00:27:28,880
moving through the business.

757
00:27:28,880 --> 00:27:30,720
Its authority comes from governance.

758
00:27:30,720 --> 00:27:32,400
If a customer order changes,

759
00:27:32,400 --> 00:27:34,480
ERP should capture that change.

760
00:27:34,480 --> 00:27:36,080
If purchasing commits to a supplier,

761
00:27:36,080 --> 00:27:37,520
ERP should record it.

762
00:27:37,520 --> 00:27:38,960
If inventory transfers,

763
00:27:38,960 --> 00:27:39,760
gets consumed,

764
00:27:39,760 --> 00:27:41,600
or enters a lot-controlled process,

765
00:27:41,600 --> 00:27:43,280
the transaction needs are home.

766
00:27:43,280 --> 00:27:46,480
That data gives APS the business context it must respect.

767
00:27:46,480 --> 00:27:48,080
APS shouldn't invent customer demand.

768
00:27:48,080 --> 00:27:50,400
It shouldn't become the owner of inventory valuation,

769
00:27:50,400 --> 00:27:52,160
supplier commitments, contract terms,

770
00:27:52,160 --> 00:27:53,680
or formal production authorization.

771
00:27:53,680 --> 00:27:56,400
It needs those facts from the systems that govern them.

772
00:27:56,400 --> 00:27:59,360
But governed business data still doesn't tell a scheduling engine

773
00:27:59,360 --> 00:28:00,960
enough to create an honest plan.

774
00:28:00,960 --> 00:28:02,560
A rooting might name a work center

775
00:28:02,560 --> 00:28:05,520
without identifying every real restriction inside it.

776
00:28:05,520 --> 00:28:07,200
Inventory might show a quantity,

777
00:28:07,200 --> 00:28:09,360
without describing exactly when that quantity

778
00:28:09,360 --> 00:28:11,200
becomes ready for a specific operation.

779
00:28:11,200 --> 00:28:14,000
Standard times may support costing well while missing the conditions

780
00:28:14,000 --> 00:28:16,800
that cause work to wait, switch, or slow down in production.

781
00:28:16,800 --> 00:28:18,640
That is the next layer of the problem.

782
00:28:18,640 --> 00:28:20,720
Before APS can schedule honestly,

783
00:28:20,720 --> 00:28:22,400
it needs detail about the constraints

784
00:28:22,400 --> 00:28:24,160
that turn an approved production route

785
00:28:24,160 --> 00:28:25,840
into work that can actually run.

786
00:28:25,840 --> 00:28:28,640
APS data, the constraint layer.

787
00:28:28,640 --> 00:28:30,160
When you're setting up APS,

788
00:28:30,160 --> 00:28:31,600
the level of detail changes

789
00:28:31,600 --> 00:28:34,640
because you're no longer tracking broad operations in a route.

790
00:28:34,640 --> 00:28:36,080
You need to understand the conditions

791
00:28:36,080 --> 00:28:38,320
that actually determine whether work can start,

792
00:28:38,320 --> 00:28:40,080
continue, and finish on the floor.

793
00:28:40,080 --> 00:28:42,640
Your ERP might say an operation belongs in machining,

794
00:28:42,640 --> 00:28:45,840
but APS needs to know which specific machine can run it.

795
00:28:45,840 --> 00:28:47,440
When that machine is available,

796
00:28:47,440 --> 00:28:50,880
what else it needs and what restrictions apply around that operation.

797
00:28:50,880 --> 00:28:52,080
Let's start with calendars.

798
00:28:52,080 --> 00:28:54,720
I don't mean Monday through Friday, 8 hours a day.

799
00:28:54,720 --> 00:28:57,120
A resource calendar has to reflect actual shifts,

800
00:28:57,120 --> 00:28:59,600
plant breaks, holidays, maintenance windows,

801
00:28:59,600 --> 00:29:01,280
and rules around overtime.

802
00:29:01,280 --> 00:29:03,120
Here's where it gets specific.

803
00:29:03,120 --> 00:29:04,960
If a machine normally runs two shifts,

804
00:29:04,960 --> 00:29:07,680
but the required operator works the first shift only,

805
00:29:07,680 --> 00:29:10,320
the machine calendar alone gives you the wrong answer.

806
00:29:10,320 --> 00:29:13,040
You're scheduling against the person, not the equipment.

807
00:29:13,040 --> 00:29:14,480
Maintenance works the same way.

808
00:29:14,480 --> 00:29:16,240
A maintenance window isn't spare capacity

809
00:29:16,240 --> 00:29:17,760
that APS can quietly consume

810
00:29:17,760 --> 00:29:19,520
because in order carries high priority.

811
00:29:19,520 --> 00:29:20,800
You can postpone that work,

812
00:29:20,800 --> 00:29:22,720
but somebody needs to approve the trade-off.

813
00:29:22,720 --> 00:29:24,880
The system should treat the window as unavailable

814
00:29:24,880 --> 00:29:27,440
until the business deliberately changes that status.

815
00:29:27,440 --> 00:29:29,680
Otherwise, you're building a schedule on assumptions

816
00:29:29,680 --> 00:29:30,880
nobody validated.

817
00:29:30,880 --> 00:29:32,640
Now capacity gets more specific.

818
00:29:32,640 --> 00:29:35,360
A work center might represent several similar machines,

819
00:29:35,360 --> 00:29:37,760
but those machines don't all carry the same capability.

820
00:29:37,760 --> 00:29:39,360
One handles a larger part size.

821
00:29:39,360 --> 00:29:42,800
Another has a particular spindle, control program, or inspection link.

822
00:29:42,800 --> 00:29:46,080
A third technically processes the part but at a slower rate.

823
00:29:46,080 --> 00:29:48,320
APS needs to model those differences

824
00:29:48,320 --> 00:29:50,560
without turning your planning model into a museum

825
00:29:50,560 --> 00:29:51,840
of every bolt in the factory.

826
00:29:51,840 --> 00:29:53,280
The goal is practical detail,

827
00:29:53,280 --> 00:29:55,040
which machines can do this operation

828
00:29:55,040 --> 00:29:56,400
how long does it take on each one?

829
00:29:56,400 --> 00:29:59,360
Which resource should APS prefer under normal conditions?

830
00:29:59,360 --> 00:30:01,120
When can it use an alternate resource

831
00:30:01,120 --> 00:30:02,640
and what changes if it does?

832
00:30:02,640 --> 00:30:04,560
People create the same kind of constraint,

833
00:30:04,560 --> 00:30:06,880
a shared operator might run two machines,

834
00:30:06,880 --> 00:30:08,320
inspect a first off part,

835
00:30:08,320 --> 00:30:11,200
or perform a setup nobody else on shift can complete.

836
00:30:11,200 --> 00:30:12,400
You can have free machine time

837
00:30:12,400 --> 00:30:15,600
and still lack the person who turns that time into actual output,

838
00:30:15,600 --> 00:30:17,280
so the schedule needs to reserve both

839
00:30:17,280 --> 00:30:18,960
when the process requires both.

840
00:30:18,960 --> 00:30:22,160
Tools, molds, and fixtures belong in the same conversation.

841
00:30:22,160 --> 00:30:24,400
A mold can't sit in two presses at once.

842
00:30:24,400 --> 00:30:27,120
A fixture might need cleaning or inspection before the next job.

843
00:30:27,120 --> 00:30:29,280
A test device might travel between lines.

844
00:30:29,280 --> 00:30:33,120
These resources often sit outside the broad ERP work center definition,

845
00:30:33,120 --> 00:30:36,000
yet they decide whether the planned job can actually run.

846
00:30:36,000 --> 00:30:38,720
Clean downtime matters just as much as run time

847
00:30:38,720 --> 00:30:40,000
in many environments.

848
00:30:40,000 --> 00:30:43,200
In food, chemical, pharmaceutical, and other process industries,

849
00:30:43,200 --> 00:30:44,960
changing from one product to another

850
00:30:44,960 --> 00:30:47,200
might require a controlled cleaning procedure

851
00:30:47,200 --> 00:30:49,840
and the duration depends on which products are involved,

852
00:30:49,840 --> 00:30:51,200
not just on the machine.

853
00:30:51,200 --> 00:30:52,720
The schedule needs to see that relationship

854
00:30:52,720 --> 00:30:55,760
because it changes how much time is actually available for production.

855
00:30:55,760 --> 00:30:58,240
Sequence rules turn all that detail into a usable schedule.

856
00:30:58,240 --> 00:31:01,120
A setup matrix tells APS how switching from one job

857
00:31:01,120 --> 00:31:03,520
to the next affects time or effort.

858
00:31:03,520 --> 00:31:05,200
Moving between products in the same family

859
00:31:05,200 --> 00:31:06,240
might cost you very little.

860
00:31:06,240 --> 00:31:08,560
Moving across families typically takes longer,

861
00:31:08,560 --> 00:31:11,040
needs a cleaning step, or requires quality approval

862
00:31:11,040 --> 00:31:12,400
before the next run starts.

863
00:31:12,400 --> 00:31:14,320
Color families are the classic example.

864
00:31:14,320 --> 00:31:16,960
A line can move from light to dark with less effort

865
00:31:16,960 --> 00:31:18,320
than from dark to light.

866
00:31:18,320 --> 00:31:20,160
Alligeant families introduce stricter rules

867
00:31:20,160 --> 00:31:22,720
where a run sequence requires cleaning and verification.

868
00:31:22,720 --> 00:31:24,160
Those aren't small preferences.

869
00:31:24,160 --> 00:31:27,280
They shape what production can safely and efficiently do.

870
00:31:27,280 --> 00:31:28,880
Campaign limits add another layer.

871
00:31:28,880 --> 00:31:31,840
You might want to group related products to reduce setup loss,

872
00:31:31,840 --> 00:31:34,080
but you also need to limit how long that campaign runs

873
00:31:34,080 --> 00:31:35,920
because another customer data approaches,

874
00:31:35,920 --> 00:31:38,320
material expires, storage fills up,

875
00:31:38,320 --> 00:31:40,480
or downstream capacity gets choked.

876
00:31:40,480 --> 00:31:42,240
APS needs the rule,

877
00:31:42,240 --> 00:31:44,880
not just the general desire to reduce changeovers.

878
00:31:44,880 --> 00:31:47,120
Qualification limits work similarly.

879
00:31:47,120 --> 00:31:48,560
A resource might appear capable

880
00:31:48,560 --> 00:31:51,360
while only certain operators can run a specific product,

881
00:31:51,360 --> 00:31:54,480
perform a release check, or handle a regulated process.

882
00:31:54,480 --> 00:31:58,160
The schedule has to respect the approved scope of those people and resources.

883
00:31:58,160 --> 00:32:01,120
Material readiness also has to go beyond stock on hand.

884
00:32:01,120 --> 00:32:02,960
A component might exist somewhere in the business

885
00:32:02,960 --> 00:32:05,600
but remain unavailable to the operation that needs it.

886
00:32:05,600 --> 00:32:07,360
It might arrive after the plan start,

887
00:32:07,360 --> 00:32:09,680
need inspections sit in a different plant area,

888
00:32:09,680 --> 00:32:12,320
or be reserved for another order that runs first.

889
00:32:12,320 --> 00:32:14,640
APS has to ask a more specific question.

890
00:32:14,640 --> 00:32:16,960
Can the required quantity reach this operation

891
00:32:16,960 --> 00:32:18,480
when the operation needs to begin?

892
00:32:18,480 --> 00:32:21,440
That question becomes more detailed when routes have alternatives.

893
00:32:21,440 --> 00:32:23,040
One product might follow a normal path

894
00:32:23,040 --> 00:32:24,240
through a preferred machine,

895
00:32:24,240 --> 00:32:27,120
but an alternate path exists for overload or equipment trouble.

896
00:32:27,120 --> 00:32:29,040
The alternate path might take longer,

897
00:32:29,040 --> 00:32:30,320
consume different tooling,

898
00:32:30,320 --> 00:32:32,080
or produce a different yield.

899
00:32:32,080 --> 00:32:35,280
Yield loss matters because an operation might need more input material

900
00:32:35,280 --> 00:32:37,200
than the final quantity suggests.

901
00:32:37,200 --> 00:32:39,760
Transfer batches matter because work can move downstream

902
00:32:39,760 --> 00:32:41,440
before the full lot finishes.

903
00:32:41,440 --> 00:32:43,840
Overlapped rules matter because the next operation

904
00:32:43,840 --> 00:32:45,920
can start after part of the batch completes

905
00:32:45,920 --> 00:32:47,520
rather than waiting for every unit.

906
00:32:47,520 --> 00:32:49,760
These details change how the schedule works.

907
00:32:49,760 --> 00:32:52,960
They tell APS how work actually flows through the plant,

908
00:32:52,960 --> 00:32:55,840
not how a broad route description suggests it should flow.

909
00:32:55,840 --> 00:32:58,000
And once you start gathering this detail,

910
00:32:58,000 --> 00:33:00,320
most teams meet the same problem early.

911
00:33:00,320 --> 00:33:02,320
Some of the planning logic lives in data

912
00:33:02,320 --> 00:33:04,800
while other parts live only in people's heads.

913
00:33:04,800 --> 00:33:07,920
Bad data does not become better because it enters APS.

914
00:33:07,920 --> 00:33:11,200
This is where a lot of APS projects meet the factory as it really works.

915
00:33:11,200 --> 00:33:13,920
The data exists, but it often describes an approved process,

916
00:33:13,920 --> 00:33:15,840
a costing model, or a release structure,

917
00:33:15,840 --> 00:33:18,400
rather than the detailed conditions that determine whether a job

918
00:33:18,400 --> 00:33:20,320
can run at a given time on a given resource.

919
00:33:20,320 --> 00:33:23,280
Take a routing with one machining operation

920
00:33:23,280 --> 00:33:25,200
and a standard time of two hours.

921
00:33:25,200 --> 00:33:28,240
That might be fine for costing and for a rough production lead time.

922
00:33:28,240 --> 00:33:31,040
But the planner knows the part only runs on one of several machines

923
00:33:31,040 --> 00:33:32,480
when a certain fixture is free,

924
00:33:32,480 --> 00:33:33,920
the first piece needs inspection,

925
00:33:33,920 --> 00:33:36,240
and a particular product version needs extra handling.

926
00:33:36,240 --> 00:33:37,920
None of that appears in the routing.

927
00:33:37,920 --> 00:33:39,520
Cycle times create the same problem.

928
00:33:39,520 --> 00:33:41,680
A standard time might come from an engineering estimate,

929
00:33:41,680 --> 00:33:44,000
an old study, or an average that includes conditions

930
00:33:44,000 --> 00:33:45,600
nobody can see from the number alone.

931
00:33:45,600 --> 00:33:48,240
Production might now run that product on newer equipment,

932
00:33:48,240 --> 00:33:49,680
use a different batch size,

933
00:33:49,680 --> 00:33:51,440
or lose time to a recurring setup

934
00:33:51,440 --> 00:33:53,280
that never entered the formal data.

935
00:33:53,280 --> 00:33:55,600
APS will schedule with the time it receives.

936
00:33:55,600 --> 00:33:57,760
It can't know that the two hour estimate translates

937
00:33:57,760 --> 00:34:00,320
to three and a half hours every Monday after a tool change.

938
00:34:00,320 --> 00:34:02,480
Generic work centers often hide the issue.

939
00:34:02,480 --> 00:34:06,560
A routing might point to assembly line A or CNC group three,

940
00:34:06,560 --> 00:34:08,480
which helps the business classify work.

941
00:34:08,480 --> 00:34:10,560
But a finite schedule needs to know

942
00:34:10,560 --> 00:34:12,160
the resources within that group,

943
00:34:12,160 --> 00:34:14,160
their actual calendars, their limits,

944
00:34:14,160 --> 00:34:16,080
and the conditions under which one machine

945
00:34:16,080 --> 00:34:17,280
can take work from another.

946
00:34:17,280 --> 00:34:20,960
Broad labels don't create a schedule people can trust.

947
00:34:20,960 --> 00:34:23,680
Setup logic often goes missing entirely.

948
00:34:23,680 --> 00:34:25,600
The planner knows that product family X

949
00:34:25,600 --> 00:34:27,280
should follow product family X,

950
00:34:27,280 --> 00:34:30,960
while family Y triggers a cleaning procedure and quality release.

951
00:34:30,960 --> 00:34:32,960
The ERP routing might only show the runtime

952
00:34:32,960 --> 00:34:35,200
because the data model never needed the relationship

953
00:34:35,200 --> 00:34:36,400
between one order and the next.

954
00:34:36,400 --> 00:34:38,560
That isn't bad behavior by the ERP team.

955
00:34:38,560 --> 00:34:40,960
It usually means the data was built for a different job.

956
00:34:40,960 --> 00:34:44,160
ERP master data typically supports product definition,

957
00:34:44,160 --> 00:34:47,120
cost, material planning, order release, and transaction control.

958
00:34:47,120 --> 00:34:48,160
Those are real requirements.

959
00:34:48,160 --> 00:34:49,680
Finite scheduling asks for more

960
00:34:49,680 --> 00:34:52,160
because it needs to simulate the operational limits of the plant,

961
00:34:52,160 --> 00:34:54,800
not just describe the approved process at a business level.

962
00:34:54,800 --> 00:34:58,000
So the question isn't whether ERP data is good or bad.

963
00:34:58,000 --> 00:35:01,680
It's whether the data fits the planning decision you want APS to support.

964
00:35:01,680 --> 00:35:05,120
If you expect APS to tell you what can run at 10 o'clock tomorrow morning,

965
00:35:05,120 --> 00:35:08,160
then resource rules, setup conditions, actual durations,

966
00:35:08,160 --> 00:35:11,200
and alternate routes need enough detail to support that decision.

967
00:35:11,200 --> 00:35:13,120
And this is where people become part of the model.

968
00:35:13,120 --> 00:35:15,200
Walk into most planning rooms and you'll find logic

969
00:35:15,200 --> 00:35:16,480
that nobody wrote down.

970
00:35:16,480 --> 00:35:19,440
A planner remembers which supplier deliveries tend to slip.

971
00:35:19,440 --> 00:35:21,680
A supervisor knows a machine can run a product,

972
00:35:21,680 --> 00:35:23,840
but only when a certain person is on shift.

973
00:35:23,840 --> 00:35:27,200
A whiteboard records a temporary restriction after a quality issue.

974
00:35:27,200 --> 00:35:28,720
Someone keeps a private spreadsheet

975
00:35:28,720 --> 00:35:30,960
because the formal system can't handle an exception

976
00:35:30,960 --> 00:35:32,080
that comes up every week.

977
00:35:32,080 --> 00:35:33,840
That knowledge keeps production moving.

978
00:35:33,840 --> 00:35:37,440
It also creates risk when the person who knows the work around is absent,

979
00:35:37,440 --> 00:35:40,960
the schedule can look correct while the floor knows it won't work.

980
00:35:40,960 --> 00:35:44,320
When APS produces a result that conflicts with that informal knowledge,

981
00:35:44,320 --> 00:35:45,840
the team may reject the schedule,

982
00:35:45,840 --> 00:35:48,000
and from their perspective, they might be right.

983
00:35:48,000 --> 00:35:50,800
Don't label that reaction as user resistance too quickly.

984
00:35:50,800 --> 00:35:54,960
More often, it's evidence that the planning model missed a real constraint.

985
00:35:54,960 --> 00:35:57,040
A decision rule, a resource limit,

986
00:35:57,040 --> 00:35:59,200
or a source of operational uncertainty.

987
00:35:59,200 --> 00:36:01,840
Sometimes the exception should stay outside the model

988
00:36:01,840 --> 00:36:03,920
because it's rare and needs human judgment.

989
00:36:03,920 --> 00:36:06,000
But if the same work around APS every week,

990
00:36:06,000 --> 00:36:07,280
it's not an exception anymore.

991
00:36:07,280 --> 00:36:08,960
It's undocumented process logic.

992
00:36:08,960 --> 00:36:10,640
Put that logic where it can be tested.

993
00:36:10,640 --> 00:36:12,480
The sensible starting point is narrow.

994
00:36:12,480 --> 00:36:14,640
Pick one constrained area, one product family,

995
00:36:14,640 --> 00:36:16,320
or one planning horizon where the pain is clear

996
00:36:16,320 --> 00:36:19,520
and the production team can compare the proposed schedule with real behavior.

997
00:36:19,520 --> 00:36:22,720
Don't begin by trying to model the whole plant in perfect detail.

998
00:36:22,720 --> 00:36:24,320
That path usually creates a long project

999
00:36:24,320 --> 00:36:26,720
and a planning model nobody has time to validate.

1000
00:36:26,720 --> 00:36:29,200
Start with enough detail to answer a real question.

1001
00:36:29,200 --> 00:36:32,960
Can this bottleneck resource meet confirmed demand over the next few weeks?

1002
00:36:32,960 --> 00:36:35,120
Given its actual shifts, setup rules,

1003
00:36:35,120 --> 00:36:36,560
and known alternatives?

1004
00:36:36,560 --> 00:36:39,200
Run the schedule against recent production history,

1005
00:36:39,200 --> 00:36:42,400
ask planners and supervisors where it predicts the wrong sequence,

1006
00:36:42,400 --> 00:36:44,640
the wrong duration or the wrong resource.

1007
00:36:44,640 --> 00:36:46,640
Then find out whether the cause lives in the data

1008
00:36:46,640 --> 00:36:48,640
in a missing business rule or in a condition

1009
00:36:48,640 --> 00:36:51,120
that changes too often for automatic scheduling.

1010
00:36:51,120 --> 00:36:54,240
That work builds trust because people can see their knowledge change the model.

1011
00:36:54,240 --> 00:36:57,200
APS doesn't clean master data by itself.

1012
00:36:57,200 --> 00:36:59,520
It exposes which parts of the data actually matter

1013
00:36:59,520 --> 00:37:01,280
when you stop planning in broad averages

1014
00:37:01,280 --> 00:37:03,760
and start planning against real constraints.

1015
00:37:03,760 --> 00:37:07,920
That can feel uncomfortable because the system puts informal assumptions out in the open.

1016
00:37:07,920 --> 00:37:09,680
But that discomfort is useful.

1017
00:37:09,680 --> 00:37:12,320
Once the model begins to describe how work actually flows,

1018
00:37:12,320 --> 00:37:14,000
the next boundary becomes clearer.

1019
00:37:14,000 --> 00:37:15,680
Planning decides what should happen.

1020
00:37:15,680 --> 00:37:18,560
Execution records what actually happens when the shift starts.

1021
00:37:18,560 --> 00:37:21,760
Where MiS fits between ERP and APS.

1022
00:37:21,760 --> 00:37:24,560
Once your planning model starts describing the real flow of work,

1023
00:37:24,560 --> 00:37:27,680
you need a way to confirm whether that work actually followed the plan.

1024
00:37:27,680 --> 00:37:28,960
That's where MES comes in.

1025
00:37:28,960 --> 00:37:30,480
MES leaves right next to the work.

1026
00:37:30,480 --> 00:37:33,760
It dispatches an operation to align, sell, or operator.

1027
00:37:33,760 --> 00:37:35,840
It records when work starts and finishes

1028
00:37:35,840 --> 00:37:39,040
how many good units came out and how much scrap happened along the way.

1029
00:37:39,040 --> 00:37:40,640
Those are execution facts.

1030
00:37:40,640 --> 00:37:43,040
Depending on the process,

1031
00:37:43,040 --> 00:37:46,000
MES might also capture downtime reasons,

1032
00:37:46,000 --> 00:37:50,320
operator actions, quality checks, machine states, and genealogy.

1033
00:37:50,320 --> 00:37:54,080
Genealogy means the traceable path of a product through production.

1034
00:37:54,080 --> 00:37:58,240
Which material lot entered the job, which machine process it,

1035
00:37:58,240 --> 00:38:00,160
which operation created the finished batch.

1036
00:38:00,160 --> 00:38:05,680
In some plants, that trail supports quality investigations or regulatory compliance.

1037
00:38:05,680 --> 00:38:07,680
In others, it helps trace a customer issue

1038
00:38:07,680 --> 00:38:10,640
without digging through paper records or relying on collective memory.

1039
00:38:10,640 --> 00:38:15,680
The distinction between MES, APS, and ERP gets clearer when you ask each system a different question.

1040
00:38:15,680 --> 00:38:16,720
ERP asks,

1041
00:38:16,720 --> 00:38:18,800
"What did the customer order? What supply do we need?

1042
00:38:18,800 --> 00:38:20,960
And what business transactions need control?"

1043
00:38:20,960 --> 00:38:24,640
APS asks, given the available resources and known constraints,

1044
00:38:24,640 --> 00:38:27,120
what sequence can meet the chosen planning goals?

1045
00:38:27,120 --> 00:38:30,400
MES asks, "What is happening now and what actually happened?"

1046
00:38:30,400 --> 00:38:32,640
That boundary sounds clean when you say it like that.

1047
00:38:32,640 --> 00:38:34,160
A real plant is rarely so neat.

1048
00:38:34,160 --> 00:38:37,200
Take a production supervisor who uses an MES dispatch list

1049
00:38:37,200 --> 00:38:39,040
as the immediate instruction for the shift.

1050
00:38:39,040 --> 00:38:43,200
That list could come from APS or the supervisor might shape it within approved rules.

1051
00:38:43,200 --> 00:38:46,080
In another plant, APS may schedule only the bottleneck,

1052
00:38:46,080 --> 00:38:49,360
while MES controls detailed dispatch across remaining operations.

1053
00:38:49,360 --> 00:38:50,560
Both approaches can work.

1054
00:38:50,560 --> 00:38:53,040
The system boundary depends on your production process,

1055
00:38:53,040 --> 00:38:56,560
the maturity of your operational data, and who owns each decision.

1056
00:38:56,560 --> 00:38:58,000
Still, the feedback loop matters.

1057
00:38:58,000 --> 00:39:01,360
APS might assume an operation starts at 8 and takes 90 minutes.

1058
00:39:01,360 --> 00:39:04,720
MES can report that the job started late because the machine needed attention.

1059
00:39:04,720 --> 00:39:07,280
It can record that actual runtime exceeded the plan,

1060
00:39:07,280 --> 00:39:09,360
that scrap reduced the usable output,

1061
00:39:09,360 --> 00:39:12,640
or that a machine state prevented the next operation from starting.

1062
00:39:12,640 --> 00:39:14,480
Those events change the planning picture.

1063
00:39:14,480 --> 00:39:16,400
If the failed job needs rework,

1064
00:39:16,400 --> 00:39:20,240
APS may need to place extra work into an already constrained schedule.

1065
00:39:20,240 --> 00:39:22,160
If yield falls below the expected level,

1066
00:39:22,160 --> 00:39:24,560
material demand and later operations may shift.

1067
00:39:24,560 --> 00:39:25,920
If a machine finishes early,

1068
00:39:25,920 --> 00:39:28,240
a resource becomes available sooner than expected.

1069
00:39:28,240 --> 00:39:32,560
A forward-looking schedule is only as credible as the execution facts that keep it current.

1070
00:39:32,560 --> 00:39:36,320
This is why MES data shouldn't become a pile of shop floor events nobody uses.

1071
00:39:36,320 --> 00:39:38,880
The data needs a defined route into planning decisions.

1072
00:39:38,880 --> 00:39:41,040
Not every event needs the same response.

1073
00:39:41,040 --> 00:39:43,360
A brief stop might matter to the line supervisor,

1074
00:39:43,360 --> 00:39:45,680
but not change the next day's schedule.

1075
00:39:45,680 --> 00:39:47,520
A long outage, an order completion,

1076
00:39:47,520 --> 00:39:48,800
a material rejection,

1077
00:39:48,800 --> 00:39:52,560
or a large yield loss may require APS to reassess what can still run.

1078
00:39:52,560 --> 00:39:54,880
The architecture needs rules for that handoff.

1079
00:39:54,880 --> 00:39:57,920
A schedule that reacts to every small signal becomes noise.

1080
00:39:57,920 --> 00:39:59,920
E-IP receives a different kind of feedback.

1081
00:39:59,920 --> 00:40:01,280
When production completes work,

1082
00:40:01,280 --> 00:40:03,280
consumes material, produces finished goods,

1083
00:40:03,280 --> 00:40:05,280
or moves stock to the next controlled state,

1084
00:40:05,280 --> 00:40:08,160
those actions often need formal confirmation in E-IP.

1085
00:40:08,160 --> 00:40:10,560
That keeps inventory, order starters, purchasing needs,

1086
00:40:10,560 --> 00:40:13,200
cost records, and financial control connected to the physical work.

1087
00:40:13,200 --> 00:40:16,080
MES captures the event at the point of execution.

1088
00:40:16,080 --> 00:40:18,880
E-IP records the government business transaction that follows.

1089
00:40:18,880 --> 00:40:22,320
That distinction matters when people try to force one system to do everything.

1090
00:40:22,320 --> 00:40:24,960
An MES doesn't need to become the financial record for the company.

1091
00:40:24,960 --> 00:40:29,600
E-IP doesn't need to receive every machine signal as though every signal changes inventory.

1092
00:40:29,600 --> 00:40:33,040
APS doesn't need to become the permanent history of every operator action.

1093
00:40:33,040 --> 00:40:34,800
Each system needs a clear job.

1094
00:40:34,800 --> 00:40:37,600
Think about a batch that reaches the end of a process.

1095
00:40:37,600 --> 00:40:40,800
MES records that production finished, records the actual quantity,

1096
00:40:40,800 --> 00:40:43,600
attaches lot information, and reports scrap.

1097
00:40:43,600 --> 00:40:46,400
APS uses that fact to update downstream availability

1098
00:40:46,400 --> 00:40:48,160
and review the remaining schedule.

1099
00:40:48,160 --> 00:40:51,120
E-IP posts the approved confirmation and inventory movement

1100
00:40:51,120 --> 00:40:53,360
so the rest of the business can act on the result.

1101
00:40:53,360 --> 00:40:55,680
One production event can matter to all three systems

1102
00:40:55,680 --> 00:40:57,520
but it means something different in each one.

1103
00:40:57,520 --> 00:40:59,680
That's why ownership needs to be explicit.

1104
00:40:59,680 --> 00:41:01,120
Who owns the production order?

1105
00:41:01,120 --> 00:41:02,880
Who owns the dispatch sequence?

1106
00:41:02,880 --> 00:41:05,120
Who decides that a schedule change becomes active?

1107
00:41:05,120 --> 00:41:08,240
Which system holds the official quantity after quality approval?

1108
00:41:08,240 --> 00:41:09,840
If those answers stay vague,

1109
00:41:09,840 --> 00:41:13,840
planners spend their day comparing dates from different places that mean different things.

1110
00:41:13,840 --> 00:41:17,680
No vendor can set those boundaries for you without understanding your process.

1111
00:41:17,680 --> 00:41:19,680
A discrete assembly plan, a batch process,

1112
00:41:19,680 --> 00:41:22,720
and a high-speed packaging line may divide the work differently.

1113
00:41:22,720 --> 00:41:24,960
Some sites run with deep MES control,

1114
00:41:24,960 --> 00:41:28,160
others rely on machine connectivity, manual confirmations,

1115
00:41:28,160 --> 00:41:30,080
and a scheduling layer that fills the gap.

1116
00:41:30,080 --> 00:41:32,720
The right architecture fits the way work runs

1117
00:41:32,720 --> 00:41:34,720
while still giving every business fact

1118
00:41:34,720 --> 00:41:36,400
and operational decision a home.

1119
00:41:36,400 --> 00:41:38,800
From there we can trace the full information loop,

1120
00:41:38,800 --> 00:41:40,560
from customer demand entering the business

1121
00:41:40,560 --> 00:41:42,720
to finished goods leaving the factory.

1122
00:41:42,720 --> 00:41:45,600
The closed planning loop, customer demand starts the loop.

1123
00:41:45,600 --> 00:41:49,200
A sales order, forecast, or confirmed customer call enters E-IP

1124
00:41:49,200 --> 00:41:51,680
and the business turns that demand into plan supply

1125
00:41:51,680 --> 00:41:54,400
through production orders, planned orders, purchase actions,

1126
00:41:54,400 --> 00:41:55,920
and material commitments.

1127
00:41:55,920 --> 00:41:58,000
That's where the plan gets its commercial meaning

1128
00:41:58,000 --> 00:41:59,840
but before the factory can act on it,

1129
00:41:59,840 --> 00:42:01,760
the order still needs a physical test.

1130
00:42:01,760 --> 00:42:04,480
It needs a root, material, capacity,

1131
00:42:04,480 --> 00:42:07,280
and a place in the sequence of work already waiting.

1132
00:42:07,280 --> 00:42:09,840
APS receives the planning inputs that let it test

1133
00:42:09,840 --> 00:42:11,280
that demand against the plant.

1134
00:42:11,280 --> 00:42:12,880
It needs the orders and their dates,

1135
00:42:12,880 --> 00:42:14,800
rootings, eligible resources,

1136
00:42:14,800 --> 00:42:17,360
shift calendars, and the current material position.

1137
00:42:17,360 --> 00:42:20,400
It also needs the planning rules that decide how work competes

1138
00:42:20,400 --> 00:42:22,400
when two jobs need the same limited resource.

1139
00:42:22,400 --> 00:42:23,440
Put those inputs together,

1140
00:42:23,440 --> 00:42:25,840
and APS can produce something production can use.

1141
00:42:25,840 --> 00:42:27,920
Sometimes that output is a detailed schedule

1142
00:42:27,920 --> 00:42:31,200
with planned, start, and finish times for each operation.

1143
00:42:31,200 --> 00:42:34,160
In other cases, especially where the floor needs room to react,

1144
00:42:34,160 --> 00:42:35,920
it's a ranked dispatch list.

1145
00:42:35,920 --> 00:42:38,560
It tells the supervisor which job should run next,

1146
00:42:38,560 --> 00:42:41,040
which should wait, and which conflict needs a decision

1147
00:42:41,040 --> 00:42:42,960
before the shift turns into guesswork.

1148
00:42:42,960 --> 00:42:44,400
Then the plan meets execution.

1149
00:42:44,400 --> 00:42:46,080
MES returns facts from production,

1150
00:42:46,080 --> 00:42:47,920
and operation starts later than planned.

1151
00:42:47,920 --> 00:42:49,200
A job completes early.

1152
00:42:49,200 --> 00:42:50,880
Scrap reduces the available quantity,

1153
00:42:50,880 --> 00:42:52,800
a quality hold stops material from moving.

1154
00:42:52,800 --> 00:42:54,400
These events tell the planning process

1155
00:42:54,400 --> 00:42:56,880
whether the assumed sequence still makes sense.

1156
00:42:56,880 --> 00:42:58,880
IoT data adds another kind of signal.

1157
00:42:58,880 --> 00:43:00,000
A machine state changes,

1158
00:43:00,000 --> 00:43:01,360
a conditional alert appears,

1159
00:43:01,360 --> 00:43:03,200
a connected asset reports it's unavailable

1160
00:43:03,200 --> 00:43:04,800
or running below its expected rate.

1161
00:43:04,800 --> 00:43:06,160
That data can matter,

1162
00:43:06,160 --> 00:43:08,160
but only when it connects to a decision.

1163
00:43:08,160 --> 00:43:10,160
A factory can produce a lot of sensor data

1164
00:43:10,160 --> 00:43:12,240
without changing a single planning action.

1165
00:43:12,240 --> 00:43:15,120
The loop needs to turn facts into controlled updates.

1166
00:43:15,120 --> 00:43:16,800
When production confirms work,

1167
00:43:16,800 --> 00:43:19,920
ERP updates order status, inventory material consumption,

1168
00:43:19,920 --> 00:43:22,240
and financial records according to business rules.

1169
00:43:22,240 --> 00:43:24,720
A completed operation may release the next step.

1170
00:43:24,720 --> 00:43:26,720
A shortage may create a new supply need.

1171
00:43:26,720 --> 00:43:28,400
A change production result can alter

1172
00:43:28,400 --> 00:43:30,400
what purchasing, warehouse, customer service,

1173
00:43:30,400 --> 00:43:31,680
and finance need to know.

1174
00:43:31,680 --> 00:43:34,400
Each system sees the same event from its own position.

1175
00:43:34,400 --> 00:43:36,640
APS sees a change in the future schedule.

1176
00:43:36,640 --> 00:43:40,080
MES sees an execution event, ERP sees a business transaction

1177
00:43:40,080 --> 00:43:41,840
or status change that needs governance.

1178
00:43:41,840 --> 00:43:44,480
That separation isn't bureaucracy for its own sake.

1179
00:43:44,480 --> 00:43:46,880
It stops one system from trying to own everything.

1180
00:43:46,880 --> 00:43:49,120
For the loop to work, the handoffs need to be clear.

1181
00:43:49,120 --> 00:43:50,480
Who owns the customer promise date?

1182
00:43:50,480 --> 00:43:52,400
Who confirms that material is usable?

1183
00:43:52,400 --> 00:43:54,800
Who decides that a revised APS schedule

1184
00:43:54,800 --> 00:43:56,400
becomes the active plan?

1185
00:43:56,400 --> 00:43:58,640
Which execution events should cause re-planning?

1186
00:43:58,640 --> 00:44:00,640
And which should remain local to the shift team?

1187
00:44:00,640 --> 00:44:02,080
If nobody answers those questions,

1188
00:44:02,080 --> 00:44:03,680
the loop breaks in quiet ways.

1189
00:44:03,680 --> 00:44:06,640
The APS schedule waits for data that arrives too late.

1190
00:44:06,640 --> 00:44:09,360
ERP dates change without the schedule being checked.

1191
00:44:09,360 --> 00:44:11,200
Supervisors work from a local priority list

1192
00:44:11,200 --> 00:44:13,680
because the released schedule no longer matches the floor.

1193
00:44:13,680 --> 00:44:15,040
Latency matters here.

1194
00:44:15,040 --> 00:44:17,760
By latency, I mean, the delay between something changing

1195
00:44:17,760 --> 00:44:19,440
in the plant and the planning process,

1196
00:44:19,440 --> 00:44:21,840
receiving a usable fact about that change.

1197
00:44:21,840 --> 00:44:23,600
If a machine stays down for most of a shift,

1198
00:44:23,600 --> 00:44:25,600
but APS learns about it the next morning,

1199
00:44:25,600 --> 00:44:28,000
the schedule has already lost its value for that decision,

1200
00:44:28,000 --> 00:44:29,520
but faster isn't always better.

1201
00:44:29,520 --> 00:44:32,080
A brief machine stop may not need an enterprise wide update.

1202
00:44:32,080 --> 00:44:34,080
A confirmed long outage probably does.

1203
00:44:34,080 --> 00:44:36,160
The right response depends on the type of event,

1204
00:44:36,160 --> 00:44:38,160
the planning horizon, and who needs to act.

1205
00:44:38,160 --> 00:44:40,720
Otherwise, you end up with a schedule that churns constantly

1206
00:44:40,720 --> 00:44:43,200
and the people on the floor stop paying attention to it.

1207
00:44:43,200 --> 00:44:45,200
Exception rules keep the loop practical.

1208
00:44:45,200 --> 00:44:47,840
They define which changes trigger a schedule review,

1209
00:44:47,840 --> 00:44:50,160
which changes can update the model automatically,

1210
00:44:50,160 --> 00:44:51,760
and which ones require a planer approval

1211
00:44:51,760 --> 00:44:53,760
before anyone sees a new dispatch order.

1212
00:44:53,760 --> 00:44:56,160
They also define what happens when the systems disagree.

1213
00:44:56,160 --> 00:44:57,040
That will happen.

1214
00:44:57,040 --> 00:45:00,240
Material may look available in ERP, while MES holds it for quality.

1215
00:45:00,240 --> 00:45:03,440
A schedule may expect a resource while maintenance takes it out of service.

1216
00:45:03,440 --> 00:45:06,000
A closed planning loop doesn't mean the factory runs itself.

1217
00:45:06,000 --> 00:45:08,800
It means the plan receives facts, tests their impact,

1218
00:45:08,800 --> 00:45:11,360
and pushes control decisions back into the work process.

1219
00:45:11,360 --> 00:45:13,520
Now, we need to be careful with one phrase that appears

1220
00:45:13,520 --> 00:45:16,160
in almost every architecture discussion, real time.

1221
00:45:16,160 --> 00:45:18,480
Real time depends on the decision.

1222
00:45:18,480 --> 00:45:20,560
Here's the real challenge with real time.

1223
00:45:20,560 --> 00:45:22,880
It sounds precise until you ask a simple question,

1224
00:45:22,880 --> 00:45:24,640
real time for which decision.

1225
00:45:24,640 --> 00:45:27,440
A machine stop needs faster awareness because the shift team

1226
00:45:27,440 --> 00:45:29,760
may need to protect people, stop feeding material,

1227
00:45:29,760 --> 00:45:32,480
call maintenance or move work before the queue builds,

1228
00:45:32,480 --> 00:45:34,320
and that can mean seconds or minutes.

1229
00:45:34,320 --> 00:45:37,440
But the same machine stop doesn't always require an instant update

1230
00:45:37,440 --> 00:45:38,880
to every enterprise plan.

1231
00:45:38,880 --> 00:45:41,120
Those are different actions with different clocks.

1232
00:45:41,120 --> 00:45:42,480
Think about the chain of events.

1233
00:45:42,480 --> 00:45:44,480
It starts when a machine state changes,

1234
00:45:44,480 --> 00:45:46,080
detected by a connected machine,

1235
00:45:46,080 --> 00:45:48,400
an MES event, or an operator report.

1236
00:45:48,400 --> 00:45:50,720
Then the planning system may recalculate the schedule

1237
00:45:50,720 --> 00:45:53,360
to see which operations now face risk.

1238
00:45:53,360 --> 00:45:55,200
After that, a planner reviews the result

1239
00:45:55,200 --> 00:45:58,160
because a proposed change may affect customer commitments,

1240
00:45:58,160 --> 00:46:01,600
overtime, maintenance, or another constrained resource.

1241
00:46:01,600 --> 00:46:03,760
Only after that review should a revised plan become

1242
00:46:03,760 --> 00:46:05,440
the released instruction for production.

1243
00:46:05,440 --> 00:46:08,320
Detection, calculation, review, and release are not one thing.

1244
00:46:08,320 --> 00:46:09,440
If you blend them together,

1245
00:46:09,440 --> 00:46:12,480
you create a system that reacts quickly, but behaves badly.

1246
00:46:12,480 --> 00:46:14,560
A short stop might trigger a recalculation

1247
00:46:14,560 --> 00:46:15,840
that moves several orders,

1248
00:46:15,840 --> 00:46:17,600
and those moves might reach the floor

1249
00:46:17,600 --> 00:46:19,760
before anyone knows whether the stop will last

1250
00:46:19,760 --> 00:46:21,360
five minutes or five hours.

1251
00:46:21,360 --> 00:46:23,680
That kind of automation doesn't create control.

1252
00:46:23,680 --> 00:46:25,280
It creates schedule noise.

1253
00:46:25,280 --> 00:46:28,160
The right refresh rhythm depends on the planning horizon.

1254
00:46:28,160 --> 00:46:30,160
A shift plan may need frequent checks

1255
00:46:30,160 --> 00:46:31,440
because the work is about to run

1256
00:46:31,440 --> 00:46:32,960
and the constraints are immediate,

1257
00:46:32,960 --> 00:46:34,560
while a near term production schedule

1258
00:46:34,560 --> 00:46:36,480
may need updates when an order completes.

1259
00:46:36,480 --> 00:46:37,920
Material becomes blocked,

1260
00:46:37,920 --> 00:46:40,960
or a resource goes down for long enough to affect the day.

1261
00:46:40,960 --> 00:46:43,120
A monthly capacity plan works differently.

1262
00:46:43,120 --> 00:46:46,080
At that level, the business is looking for broader pressure,

1263
00:46:46,080 --> 00:46:47,920
expected demand, supply risk,

1264
00:46:47,920 --> 00:46:49,840
plan maintenance, and weather capacity

1265
00:46:49,840 --> 00:46:51,520
will support the sales plan.

1266
00:46:51,520 --> 00:46:54,560
It doesn't need a response every time a sensor sees a short speed loss

1267
00:46:54,560 --> 00:46:56,720
or align pauses for a normal adjustment.

1268
00:46:56,720 --> 00:46:58,960
Sending those signals into a monthly planning process

1269
00:46:58,960 --> 00:47:00,240
would be like calling a board meeting

1270
00:47:00,240 --> 00:47:02,560
because a forklift stopped for a battery change.

1271
00:47:02,560 --> 00:47:03,840
You need decision windows.

1272
00:47:03,840 --> 00:47:05,840
For safety and immediate operations,

1273
00:47:05,840 --> 00:47:07,200
the window may be seconds.

1274
00:47:07,200 --> 00:47:09,440
A supervisor deciding whether to re-sequence work

1275
00:47:09,440 --> 00:47:11,440
during the shift may need it in minutes.

1276
00:47:11,440 --> 00:47:13,120
A planner reviewing the next few days

1277
00:47:13,120 --> 00:47:15,760
may work with one or more scheduled review points during the day,

1278
00:47:15,760 --> 00:47:18,400
and for purchasing, sales, and capacity planning,

1279
00:47:18,400 --> 00:47:21,120
the useful window may stretch across days or weeks.

1280
00:47:21,120 --> 00:47:23,840
The window should follow the decision, not the technology.

1281
00:47:23,840 --> 00:47:26,320
This also affects how you connect OT and IT.

1282
00:47:26,320 --> 00:47:29,360
Operational technology or OT produces a stream of events

1283
00:47:29,360 --> 00:47:31,920
from machines, sensors, and production systems.

1284
00:47:31,920 --> 00:47:33,520
Some events belong in local control,

1285
00:47:33,520 --> 00:47:34,880
some should reach MES,

1286
00:47:34,880 --> 00:47:37,040
a smaller set should trigger a planning review,

1287
00:47:37,040 --> 00:47:38,560
and only selected events should create

1288
00:47:38,560 --> 00:47:40,720
or change a governed ERP transaction.

1289
00:47:40,720 --> 00:47:42,720
Not every sensor signal means stock changed,

1290
00:47:42,720 --> 00:47:44,720
not every stop means a customer day changed,

1291
00:47:44,720 --> 00:47:47,440
and not every machine alarm needs a new production order status.

1292
00:47:47,440 --> 00:47:50,480
ERP needs business facts that people can audit and act on.

1293
00:47:50,480 --> 00:47:53,200
If you push every low-level event into ERP,

1294
00:47:53,200 --> 00:47:54,960
you create volume without meaning,

1295
00:47:54,960 --> 00:47:57,120
and you risk turning a control transaction system

1296
00:47:57,120 --> 00:47:58,880
into a very expensive event log,

1297
00:47:58,880 --> 00:48:01,760
that doesn't mean ERP should stay disconnected from the floor.

1298
00:48:01,760 --> 00:48:03,680
It means the integration needs intent.

1299
00:48:03,680 --> 00:48:05,760
A completed operation, a confirmed quantity,

1300
00:48:05,760 --> 00:48:08,560
a quality hold, a long outage with schedule impact

1301
00:48:08,560 --> 00:48:10,080
or material consumption event

1302
00:48:10,080 --> 00:48:12,960
may need to flow into ERP through defined rules.

1303
00:48:12,960 --> 00:48:14,720
A momentary temperature fluctuation

1304
00:48:14,720 --> 00:48:16,320
may matter to engineering or maintenance

1305
00:48:16,320 --> 00:48:18,400
without changing a business transaction at all.

1306
00:48:18,400 --> 00:48:20,160
In practical terms, define the trigger

1307
00:48:20,160 --> 00:48:21,760
before you build the interface.

1308
00:48:21,760 --> 00:48:24,000
Ask who needs to know what decision changes,

1309
00:48:24,000 --> 00:48:25,280
how fast must they act,

1310
00:48:25,280 --> 00:48:27,040
and does this event need a plan as approval

1311
00:48:27,040 --> 00:48:29,040
before it changes the released schedule?

1312
00:48:29,040 --> 00:48:30,160
Those questions sound basic,

1313
00:48:30,160 --> 00:48:32,000
but they stop a lot of bad architecture.

1314
00:48:32,000 --> 00:48:33,120
Teams often start with,

1315
00:48:33,120 --> 00:48:34,560
can we stream this data?

1316
00:48:34,560 --> 00:48:35,920
The better question is,

1317
00:48:35,920 --> 00:48:38,400
what will somebody do differently when this data arrives?

1318
00:48:38,400 --> 00:48:40,720
Once you frame real time

1319
00:48:40,720 --> 00:48:43,120
around decisions the architecture becomes calmer,

1320
00:48:43,120 --> 00:48:45,360
events reach the people and systems that need them,

1321
00:48:45,360 --> 00:48:47,520
planning updates when the impact warrants it,

1322
00:48:47,520 --> 00:48:49,520
and ERP records control changes

1323
00:48:49,520 --> 00:48:51,840
instead of every twitch from the factory floor,

1324
00:48:51,840 --> 00:48:53,200
which leads to a procurement question

1325
00:48:53,200 --> 00:48:54,640
people often frame badly.

1326
00:48:54,640 --> 00:48:56,960
Should you choose APS or ERP?

1327
00:48:56,960 --> 00:49:00,720
APS versus ERP is usually the wrong procurement question.

1328
00:49:00,720 --> 00:49:02,800
Should we choose APS or ERP?

1329
00:49:02,800 --> 00:49:04,320
Sounds like a clean procurement question,

1330
00:49:04,320 --> 00:49:05,440
but it usually isn't.

1331
00:49:05,440 --> 00:49:08,000
It assumes both systems compete for the same job

1332
00:49:08,000 --> 00:49:09,520
when they normally deal with different parts

1333
00:49:09,520 --> 00:49:10,960
of one planning chain.

1334
00:49:10,960 --> 00:49:13,600
ERP gives the business a controlled view of demand,

1335
00:49:13,600 --> 00:49:16,160
supply orders, inventory, purchasing,

1336
00:49:16,160 --> 00:49:17,680
and financial responsibility.

1337
00:49:17,680 --> 00:49:20,000
APS takes the work that the business wants to do

1338
00:49:20,000 --> 00:49:21,600
and tests it against the operating limits

1339
00:49:21,600 --> 00:49:23,360
that decide whether the plan can happen.

1340
00:49:23,360 --> 00:49:26,160
One creates and governs the demand signal,

1341
00:49:26,160 --> 00:49:27,760
while the other challenges that signal

1342
00:49:27,760 --> 00:49:30,560
against time, capacity, sequence, and constraints.

1343
00:49:30,560 --> 00:49:31,600
You need both roles,

1344
00:49:31,600 --> 00:49:34,160
even if one system delivers more of the planning function

1345
00:49:34,160 --> 00:49:35,200
in a given plant.

1346
00:49:35,200 --> 00:49:37,760
An ERP-only approach can work well in the right setting.

1347
00:49:37,760 --> 00:49:39,680
Picture a plant with stable product routes,

1348
00:49:39,680 --> 00:49:42,080
a limited range of products, predictable demand,

1349
00:49:42,080 --> 00:49:44,880
and enough spare capacity that a small change in one order

1350
00:49:44,880 --> 00:49:47,120
doesn't force the whole schedule to move.

1351
00:49:47,120 --> 00:49:50,160
In that environment, standard lead times may be close enough

1352
00:49:50,160 --> 00:49:52,560
and planners may know the few constraints that matter,

1353
00:49:52,560 --> 00:49:54,400
handling exceptions with direct contact

1354
00:49:54,400 --> 00:49:56,720
between production, purchasing, and sales.

1355
00:49:56,720 --> 00:49:59,200
The cost and effort of adding another scheduling layer

1356
00:49:59,200 --> 00:50:01,200
may outweigh the planning problem.

1357
00:50:01,200 --> 00:50:02,640
That isn't a failure of maturity

1358
00:50:02,640 --> 00:50:05,520
and it's a sensible fit between the process and the tools.

1359
00:50:05,520 --> 00:50:08,720
The pressure rises when the factory loses that simplicity.

1360
00:50:08,720 --> 00:50:11,280
Shared machines create competition between orders.

1361
00:50:11,280 --> 00:50:14,160
Variable lead times make all assumptions less useful

1362
00:50:14,160 --> 00:50:17,840
and complex setups mean the sequence changes the available capacity.

1363
00:50:17,840 --> 00:50:21,840
A shortage, late delivery, quality hold, machine issue, or customer request

1364
00:50:21,840 --> 00:50:23,760
can turn a plan that looked reasonable yesterday

1365
00:50:23,760 --> 00:50:25,760
into a set of conflicting choices today.

1366
00:50:25,760 --> 00:50:28,960
Then, planners need more than a date and a work order status.

1367
00:50:28,960 --> 00:50:31,680
They need to test consequences before they release work

1368
00:50:31,680 --> 00:50:33,760
or promise a customer a new date.

1369
00:50:33,760 --> 00:50:35,920
Can this order move to another resource?

1370
00:50:35,920 --> 00:50:37,440
What work moves if it does?

1371
00:50:37,440 --> 00:50:40,080
Does the alternate route require different labor or tooling

1372
00:50:40,080 --> 00:50:42,080
will a change that saves one customer order

1373
00:50:42,080 --> 00:50:44,000
create late orders somewhere else?

1374
00:50:44,000 --> 00:50:47,040
An ERP can hold many of the facts needed to ask those questions

1375
00:50:47,040 --> 00:50:48,880
but it doesn't always contain the scheduling logic

1376
00:50:48,880 --> 00:50:50,640
or constraint detail needed to answer them

1377
00:50:50,640 --> 00:50:52,000
at the level the factory needs.

1378
00:50:52,000 --> 00:50:53,600
That is when APS becomes interesting,

1379
00:50:53,600 --> 00:50:55,360
not because the term sounds more advanced

1380
00:50:55,360 --> 00:50:58,240
or because a vendor says every manufacturer needs an optimizer.

1381
00:50:58,240 --> 00:51:00,720
APS earns its place when the cost of planning

1382
00:51:00,720 --> 00:51:02,720
from broad assumptions becomes higher

1383
00:51:02,720 --> 00:51:04,880
than the cost of modeling the real constraints.

1384
00:51:04,880 --> 00:51:06,880
There is an equal mistake on the other side.

1385
00:51:06,880 --> 00:51:09,520
Some teams see APS as the place where the schedule happens

1386
00:51:09,520 --> 00:51:12,320
then assume it can become the center of the whole planning process.

1387
00:51:12,320 --> 00:51:15,040
But APS without ERP can leave serious gaps.

1388
00:51:15,040 --> 00:51:16,720
Where does confirmed demand come from?

1389
00:51:16,720 --> 00:51:18,480
Who owns the approved bill of materials

1390
00:51:18,480 --> 00:51:22,880
which inventory is available, blocked, committed, or financially controlled?

1391
00:51:22,880 --> 00:51:25,120
How do purchasing commitments, production conformations,

1392
00:51:25,120 --> 00:51:27,600
and customer changes remain governed across the business?

1393
00:51:27,600 --> 00:51:29,600
A schedule without those business facts

1394
00:51:29,600 --> 00:51:31,360
may look practical for the next shift

1395
00:51:31,360 --> 00:51:34,160
while drifting away from what the company has actually sold,

1396
00:51:34,160 --> 00:51:36,000
bought, approved, and promised.

1397
00:51:36,000 --> 00:51:37,760
APS shouldn't become a shadow ERP

1398
00:51:37,760 --> 00:51:40,480
because that path creates duplicate orders, duplicate material views,

1399
00:51:40,480 --> 00:51:43,120
and arguments about which date or quantity people should trust.

1400
00:51:43,120 --> 00:51:44,800
You haven't removed complexity.

1401
00:51:44,800 --> 00:51:46,960
You've moved it into interfaces and manual checks

1402
00:51:46,960 --> 00:51:49,200
where it tends to return at inconvenient times.

1403
00:51:49,200 --> 00:51:51,760
So don't begin the discussion with a product category.

1404
00:51:51,760 --> 00:51:53,360
Start with the planning failure.

1405
00:51:53,360 --> 00:51:55,120
Where does the current process break down?

1406
00:51:55,120 --> 00:51:57,440
Does material appear available in the plan

1407
00:51:57,440 --> 00:51:59,440
but fail to arrive at the point of use?

1408
00:51:59,440 --> 00:52:03,040
Do customer dates change because nobody tested finite capacity?

1409
00:52:03,040 --> 00:52:06,400
Do planners spend hours rebuilding a sequence after every disruption?

1410
00:52:06,400 --> 00:52:08,960
Does production receive work that cannot start?

1411
00:52:08,960 --> 00:52:11,440
While urgent orders remain hidden in a long list?

1412
00:52:11,440 --> 00:52:13,760
Those questions point to the missing information.

1413
00:52:13,760 --> 00:52:16,400
Sometimes the missing piece is more accurately timed data.

1414
00:52:16,400 --> 00:52:18,480
Other times its machine-level capacity,

1415
00:52:18,480 --> 00:52:20,240
setup rules, qualified labor,

1416
00:52:20,240 --> 00:52:22,880
or a clearer process for who can change priorities.

1417
00:52:22,880 --> 00:52:25,440
And sometimes the real issue sits in master data

1418
00:52:25,440 --> 00:52:27,920
or in the hand-off between planning and execution,

1419
00:52:27,920 --> 00:52:29,840
not in the absence of an APS product.

1420
00:52:29,840 --> 00:52:32,080
That distinction protects you from buying software

1421
00:52:32,080 --> 00:52:34,480
to solve an operating problem nobody has defined.

1422
00:52:34,480 --> 00:52:37,040
A sensible architecture begins with ownership.

1423
00:52:37,040 --> 00:52:39,520
ERP owns governed business transactions.

1424
00:52:39,520 --> 00:52:42,000
APS owns the constrained planning view

1425
00:52:42,000 --> 00:52:43,760
where that capability exists.

1426
00:52:43,760 --> 00:52:45,760
And the integration needs to carry the facts

1427
00:52:45,760 --> 00:52:48,080
each system requires without pretending

1428
00:52:48,080 --> 00:52:50,080
that either system should own every decision.

1429
00:52:50,080 --> 00:52:52,640
Then you can ask a much better question

1430
00:52:52,640 --> 00:52:55,280
which planning decisions need a constraint-aware answer.

1431
00:52:55,280 --> 00:52:58,160
And what data must exist before that answer deserves trust.

1432
00:52:58,160 --> 00:52:59,520
There is still one complication

1433
00:52:59,520 --> 00:53:03,200
because many ERP platforms include planning and scheduling functions.

1434
00:53:03,200 --> 00:53:05,760
So the next question isn't simply whether you need APS.

1435
00:53:05,760 --> 00:53:07,600
It's whether the scheduling capability

1436
00:53:07,600 --> 00:53:10,720
already inside your ERP can handle the problem you actually have.

1437
00:53:10,720 --> 00:53:13,360
When ERP scheduling may be enough,

1438
00:53:13,360 --> 00:53:15,360
here's the thing about planning complexity.

1439
00:53:15,360 --> 00:53:17,360
Just because your plant makes complicated stuff

1440
00:53:17,360 --> 00:53:20,480
doesn't mean you automatically need dedicated APS software.

1441
00:53:20,480 --> 00:53:22,000
A lot of factories have a planning problem

1442
00:53:22,000 --> 00:53:23,440
that ERP scheduling,

1443
00:53:23,440 --> 00:53:26,640
combined with some solid routines, can handle just fine.

1444
00:53:26,640 --> 00:53:28,400
Picture a plant where roots are stable,

1445
00:53:28,400 --> 00:53:30,000
the product ranges small

1446
00:53:30,000 --> 00:53:32,080
and demand stays pretty steady week to week.

1447
00:53:32,080 --> 00:53:34,720
The same products hit the same resources every time

1448
00:53:34,720 --> 00:53:37,280
and most orders follow a familiar path through production.

1449
00:53:37,280 --> 00:53:38,400
Capacity still matters,

1450
00:53:38,400 --> 00:53:40,240
but there's rarely a traffic jam at the bottleneck

1451
00:53:40,240 --> 00:53:41,600
that shifts the planning burden.

1452
00:53:41,600 --> 00:53:43,280
If the plant keeps some capacity headroom,

1453
00:53:43,280 --> 00:53:44,800
the plan doesn't fall apart every time

1454
00:53:44,800 --> 00:53:46,080
an order shifts by a few hours.

1455
00:53:46,080 --> 00:53:47,760
The planner can lean on lead times,

1456
00:53:47,760 --> 00:53:48,800
work center loads,

1457
00:53:48,800 --> 00:53:51,840
and a short term view of open orders to keep work moving.

1458
00:53:51,840 --> 00:53:53,440
Production knows the normal rhythm

1459
00:53:53,440 --> 00:53:54,720
and exceptions stay exceptions

1460
00:53:54,720 --> 00:53:56,240
instead of becoming the daily reality.

1461
00:53:56,240 --> 00:53:57,280
So in that situation,

1462
00:53:57,280 --> 00:53:59,440
rough cut capacity planning might be all you need.

1463
00:53:59,440 --> 00:54:00,960
It compares expected demand

1464
00:54:00,960 --> 00:54:03,040
with available capacity at a broad level.

1465
00:54:03,040 --> 00:54:04,880
It won't give you every exact start time,

1466
00:54:04,880 --> 00:54:07,280
but it will tell you whether a resource group has more work

1467
00:54:07,280 --> 00:54:09,120
than it can handle over the next period.

1468
00:54:09,120 --> 00:54:10,000
For a lot of plants,

1469
00:54:10,000 --> 00:54:12,160
that's really the only decision they need to make.

1470
00:54:12,160 --> 00:54:14,080
Simple finite scheduling can also do the job

1471
00:54:14,080 --> 00:54:15,600
for near-term control.

1472
00:54:15,600 --> 00:54:17,840
An ERP scheduling function can handle

1473
00:54:17,840 --> 00:54:20,400
a limited set of resources against working calendars,

1474
00:54:20,400 --> 00:54:22,160
account for basic capacity limits

1475
00:54:22,160 --> 00:54:24,160
and show planners where orders overlap.

1476
00:54:24,160 --> 00:54:25,520
If set up logic is simple,

1477
00:54:25,520 --> 00:54:26,960
alternate routes are rare,

1478
00:54:26,960 --> 00:54:29,360
and labor constraints stay consistent day to day.

1479
00:54:29,360 --> 00:54:30,880
That's often enough control.

1480
00:54:30,880 --> 00:54:32,160
Enough is the word that matters.

1481
00:54:32,160 --> 00:54:33,520
Nobody gets a trophy for building

1482
00:54:33,520 --> 00:54:35,520
the most complex planning architecture.

1483
00:54:35,520 --> 00:54:37,760
If a simpler process gives planners a credible answer,

1484
00:54:37,760 --> 00:54:39,200
and production can actually follow it

1485
00:54:39,200 --> 00:54:40,800
without a daily manual rescue,

1486
00:54:40,800 --> 00:54:42,240
then adding a separate system

1487
00:54:42,240 --> 00:54:44,320
might create more burden than benefit.

1488
00:54:44,320 --> 00:54:47,120
Strong planner routines can close some of that gap as well.

1489
00:54:47,120 --> 00:54:49,920
An experienced planner knows the handful of constraints

1490
00:54:49,920 --> 00:54:51,200
that don't show up in the system

1491
00:54:51,200 --> 00:54:52,800
and deals with them through regular chats

1492
00:54:52,800 --> 00:54:55,520
with supervisors, purchasing, and customer service.

1493
00:54:55,520 --> 00:54:56,800
A daily planning meeting can work

1494
00:54:56,800 --> 00:54:58,160
when the plant is manageable,

1495
00:54:58,160 --> 00:54:59,760
the exceptions stay visible,

1496
00:54:59,760 --> 00:55:00,960
and decisions don't require

1497
00:55:00,960 --> 00:55:03,840
rebuilding hundreds of linked operations under pressure.

1498
00:55:03,840 --> 00:55:06,000
That's not a reason to dismiss the planner's knowledge.

1499
00:55:06,000 --> 00:55:07,360
It's part of how the factory runs,

1500
00:55:07,360 --> 00:55:09,600
but you need to test whether that knowledge scales.

1501
00:55:09,600 --> 00:55:11,120
If the planner takes a vacation,

1502
00:55:11,120 --> 00:55:13,280
switches roles or just faces a week,

1503
00:55:13,280 --> 00:55:15,520
where demand and disruption hit at the same time,

1504
00:55:15,520 --> 00:55:17,200
does the process still produce a plan,

1505
00:55:17,200 --> 00:55:18,240
people can trust?

1506
00:55:18,240 --> 00:55:19,600
If the answer is yes,

1507
00:55:19,600 --> 00:55:22,320
then the planning problem probably still fits inside ERP

1508
00:55:22,320 --> 00:55:24,480
with a well-run operating routine,

1509
00:55:24,480 --> 00:55:26,480
the cost of another system matters as well.

1510
00:55:26,480 --> 00:55:28,720
A dedicated APS product needs data prep,

1511
00:55:28,720 --> 00:55:30,480
integration, training, process, design,

1512
00:55:30,480 --> 00:55:31,920
support, and ongoing ownership.

1513
00:55:31,920 --> 00:55:33,440
It might require changes to routes,

1514
00:55:33,440 --> 00:55:34,480
work centers, calendars,

1515
00:55:34,480 --> 00:55:36,480
and how production reports actual progress.

1516
00:55:36,480 --> 00:55:37,680
Those changes can pay off,

1517
00:55:37,680 --> 00:55:39,120
but they aren't free,

1518
00:55:39,120 --> 00:55:40,640
and they don't get any easier

1519
00:55:40,640 --> 00:55:42,800
just because the sales demo made them look perfect.

1520
00:55:42,800 --> 00:55:44,000
You've got to weigh that effort

1521
00:55:44,000 --> 00:55:46,080
against the pain you're actually trying to remove.

1522
00:55:46,080 --> 00:55:47,440
Ask yourself,

1523
00:55:47,440 --> 00:55:51,360
are missed dates frequent because the schedule can't test capacity?

1524
00:55:51,360 --> 00:55:54,160
Are planners spending a big part of each day rebuilding the plan?

1525
00:55:54,160 --> 00:55:56,160
Does production often start work

1526
00:55:56,160 --> 00:55:58,800
that later waits for a shared resource or material?

1527
00:55:58,800 --> 00:56:00,640
Or does the existing process mostly work

1528
00:56:00,640 --> 00:56:03,200
with occasional manual adjustments the team can handle?

1529
00:56:03,200 --> 00:56:04,560
Those are very different situations.

1530
00:56:04,560 --> 00:56:06,320
Vendor checklists won't answer this for you.

1531
00:56:06,320 --> 00:56:09,200
Most systems can claim capacity planning,

1532
00:56:09,200 --> 00:56:10,720
scheduling, alerts,

1533
00:56:10,720 --> 00:56:12,560
what-if scenarios and integrations.

1534
00:56:12,560 --> 00:56:15,520
The real test comes from your own recent production cases.

1535
00:56:15,520 --> 00:56:17,840
Take a normal week, a late customer request,

1536
00:56:17,840 --> 00:56:20,640
or a resource outage that planners remember clearly.

1537
00:56:20,640 --> 00:56:22,480
Then ask the ERP scheduling capability

1538
00:56:22,480 --> 00:56:25,440
to model those cases using your routes, calendars, and rules.

1539
00:56:25,440 --> 00:56:26,560
Can it explain the result?

1540
00:56:26,560 --> 00:56:28,080
Can a planner change an assumption

1541
00:56:28,080 --> 00:56:30,240
without rebuilding the plan by hand?

1542
00:56:30,240 --> 00:56:31,840
Can production act on the output?

1543
00:56:31,840 --> 00:56:32,960
That gives you real evidence.

1544
00:56:32,960 --> 00:56:34,160
If the answer is yes,

1545
00:56:34,160 --> 00:56:36,080
you may not need a separate APS system yet.

1546
00:56:36,080 --> 00:56:37,440
Focus on improving the process

1547
00:56:37,440 --> 00:56:39,440
and tightening the data that affects planning

1548
00:56:39,440 --> 00:56:41,600
and avoid adding technology for a problem

1549
00:56:41,600 --> 00:56:42,960
you don't actually have.

1550
00:56:42,960 --> 00:56:45,200
But if the ERP function keeps exposing conflicts

1551
00:56:45,200 --> 00:56:47,200
it can't model or test fast enough

1552
00:56:47,200 --> 00:56:48,960
then the planning problem has shifted.

1553
00:56:48,960 --> 00:56:51,760
That's when dedicated APS starts making real sense.

1554
00:56:51,760 --> 00:56:54,240
When dedicated APS becomes hard to avoid,

1555
00:56:54,240 --> 00:56:56,800
a dedicated APS system starts earning attention

1556
00:56:56,800 --> 00:56:59,920
when the planner is no longer managing a simple queue of work.

1557
00:56:59,920 --> 00:57:02,080
The hard cases begin when multiple orders

1558
00:57:02,080 --> 00:57:04,240
compete for the same limited resources

1559
00:57:04,240 --> 00:57:07,040
and every decision shifts the outcome for something else.

1560
00:57:07,040 --> 00:57:09,520
Picture a factory where several product lines

1561
00:57:09,520 --> 00:57:12,160
all need the same heat treatment furnace before they can ship.

1562
00:57:12,160 --> 00:57:14,240
The furnace has fixed available hours.

1563
00:57:14,240 --> 00:57:16,800
One order is late, another carries a customer commitment

1564
00:57:16,800 --> 00:57:19,600
and a third needs to run while a certain material lot

1565
00:57:19,600 --> 00:57:21,440
stays within its approved use window.

1566
00:57:21,440 --> 00:57:23,440
The question isn't whether the furnace has work,

1567
00:57:23,440 --> 00:57:24,320
it always has work.

1568
00:57:24,320 --> 00:57:27,120
The question is which job gets the next available slot

1569
00:57:27,120 --> 00:57:29,040
and what the business gives up with that choice.

1570
00:57:29,040 --> 00:57:31,520
That's a finite scheduling problem, pure and simple.

1571
00:57:31,520 --> 00:57:33,600
The complexity grows when the sequence

1572
00:57:33,600 --> 00:57:35,440
changes the available capacity.

1573
00:57:35,440 --> 00:57:37,440
A machine might need very little preparation

1574
00:57:37,440 --> 00:57:38,720
between related products

1575
00:57:38,720 --> 00:57:41,680
then require a long setup for the next job from another family.

1576
00:57:41,680 --> 00:57:43,520
One alternate machine could take the work

1577
00:57:43,520 --> 00:57:44,800
but at a slower rate.

1578
00:57:44,800 --> 00:57:46,880
Another might run it only with a qualified operator

1579
00:57:46,880 --> 00:57:48,720
who also supports a different bottle lake.

1580
00:57:48,720 --> 00:57:51,760
Available time on a machine is just part of the equation.

1581
00:57:51,760 --> 00:57:54,160
Material timing adds another layer of pressure.

1582
00:57:54,160 --> 00:57:56,160
A component might arrive later in the day

1583
00:57:56,160 --> 00:57:57,680
while another order can start now

1584
00:57:57,680 --> 00:57:59,360
but needs a fixture that's tied up elsewhere.

1585
00:57:59,360 --> 00:58:01,840
The planner can't treat those conditions as separate lists.

1586
00:58:01,840 --> 00:58:03,040
They interact.

1587
00:58:03,040 --> 00:58:04,880
Moving in order to solve a material delay

1588
00:58:04,880 --> 00:58:06,480
can consume the only resource needed

1589
00:58:06,480 --> 00:58:08,960
for a higher priority order later in the shift.

1590
00:58:08,960 --> 00:58:10,720
Manual planning can still handle this

1591
00:58:10,720 --> 00:58:12,640
right up to the point where it can't.

1592
00:58:12,640 --> 00:58:15,120
Most experienced planners can make smart calls from memory.

1593
00:58:15,120 --> 00:58:18,000
They know which product runs badly on an alternate machine

1594
00:58:18,000 --> 00:58:20,000
which supervisor can solve a setup issue

1595
00:58:20,000 --> 00:58:21,920
and which customer promise has room for discussion.

1596
00:58:21,920 --> 00:58:25,600
But frequent disruptions turn that knowledge into a burden

1597
00:58:25,600 --> 00:58:27,360
because every change forces the planner

1598
00:58:27,360 --> 00:58:29,840
to recalculate a growing chain of consequences.

1599
00:58:29,840 --> 00:58:31,920
A late supplier delivery, a tool failure,

1600
00:58:31,920 --> 00:58:34,640
a quality hold on a batch or a rush order from sales.

1601
00:58:34,640 --> 00:58:36,720
None of these events need to be dramatic on their own

1602
00:58:36,720 --> 00:58:38,320
but together they force the planner

1603
00:58:38,320 --> 00:58:40,800
to test more options than a whiteboard, a spreadsheet

1604
00:58:40,800 --> 00:58:43,520
and a few phone calls can support in the time available.

1605
00:58:43,520 --> 00:58:45,600
Speed matters but not for its own sake.

1606
00:58:45,600 --> 00:58:47,600
The reason to test scenarios quickly

1607
00:58:47,600 --> 00:58:50,720
is that customer promise dates carry real commercial weight.

1608
00:58:50,720 --> 00:58:53,280
Once sales tells a customer an order can ship on Friday

1609
00:58:53,280 --> 00:58:55,200
the plant has made a commitment.

1610
00:58:55,200 --> 00:58:57,680
Changing that later can affect trust, expedite costs,

1611
00:58:57,680 --> 00:59:00,400
downstream logistics, or the customer's own production plan.

1612
00:59:00,400 --> 00:59:02,000
Before anyone changes a promise date

1613
00:59:02,000 --> 00:59:04,000
they need to understand the downstream impact

1614
00:59:04,000 --> 00:59:06,640
can the plant protect this order by moving another job?

1615
00:59:06,640 --> 00:59:08,400
Does that move create a later conflict

1616
00:59:08,400 --> 00:59:10,320
that assembly inspection or packaging?

1617
00:59:10,320 --> 00:59:11,680
Does the schedule need overtime

1618
00:59:11,680 --> 00:59:13,040
and is the right labor available?

1619
00:59:13,040 --> 00:59:15,600
And is the proposed recovery plan actually physically possible

1620
00:59:15,600 --> 00:59:19,040
or does it only look possible because the model ignores a shared constraint?

1621
00:59:19,040 --> 00:59:20,800
A dedicated APS engine helps here

1622
00:59:20,800 --> 00:59:22,720
because it can evaluate these linked conditions

1623
00:59:22,720 --> 00:59:24,320
repeatedly and consistently.

1624
00:59:24,320 --> 00:59:26,240
It can test the scenario without pretending

1625
00:59:26,240 --> 00:59:27,920
the first answer is the only answer.

1626
00:59:27,920 --> 00:59:30,880
The planner can compare options, keep the current sequence,

1627
00:59:30,880 --> 00:59:33,440
move the urgent order forward, use an alternate resource

1628
00:59:33,440 --> 00:59:35,600
or delay a lower priority order.

1629
00:59:35,600 --> 00:59:37,280
Each option needs an explanation,

1630
00:59:37,280 --> 00:59:39,120
not just a new date on a screen.

1631
00:59:39,120 --> 00:59:40,720
The need becomes even clearer

1632
00:59:40,720 --> 00:59:42,560
when you're dealing with more than one plant.

1633
00:59:42,560 --> 00:59:43,920
A company might have two sites

1634
00:59:43,920 --> 00:59:45,360
that can produce related products,

1635
00:59:45,360 --> 00:59:46,560
share a scarce component,

1636
00:59:46,560 --> 00:59:48,320
or depend on the same specialist team.

1637
00:59:48,320 --> 00:59:50,720
A local plan might look workable at each site

1638
00:59:50,720 --> 00:59:53,600
while the combined plan ends up consuming the same material twice

1639
00:59:53,600 --> 00:59:55,520
or overloading the shared resource.

1640
00:59:55,520 --> 00:59:57,520
That's where planning stops at the factory gate.

1641
00:59:57,520 --> 01:00:00,560
Capacity decisions might involve where to build an order,

1642
01:00:00,560 --> 01:00:01,920
where to hold inventory,

1643
01:00:01,920 --> 01:00:03,840
and how to protect the customer commitment

1644
01:00:03,840 --> 01:00:06,160
when one site loses output.

1645
01:00:06,160 --> 01:00:09,280
ERP can hold the commercial structure around those choices.

1646
01:00:09,280 --> 01:00:12,320
A dedicated APS layer can test the constrained production parts

1647
01:00:12,320 --> 01:00:13,440
across the network,

1648
01:00:13,440 --> 01:00:16,160
as long as the organization has modeled the real limits

1649
01:00:16,160 --> 01:00:18,160
and agreed who can make the trade-offs.

1650
01:00:18,160 --> 01:00:20,000
No software removes those trade-offs.

1651
01:00:20,000 --> 01:00:22,400
It just makes them visible before people commit to them

1652
01:00:22,400 --> 01:00:24,800
and that leads to a distinction worth keeping clear.

1653
01:00:24,800 --> 01:00:28,000
An APS engine can calculate a schedule under the rules it receives

1654
01:00:28,000 --> 01:00:30,000
but a planner still decides whether those rules

1655
01:00:30,000 --> 01:00:31,680
represent the right business judgment.

1656
01:00:31,680 --> 01:00:33,680
APS does not replace the planner.

1657
01:00:33,680 --> 01:00:36,080
Here's the thing about dedicated scheduling software

1658
01:00:36,080 --> 01:00:38,640
an APS system can calculate way more schedule options

1659
01:00:38,640 --> 01:00:40,560
than any planner could rebuild by hand

1660
01:00:40,560 --> 01:00:42,080
and that's genuinely useful.

1661
01:00:42,080 --> 01:00:44,400
But it doesn't mean the system should take over the planner's job

1662
01:00:44,400 --> 01:00:46,080
because a lot of the real decisions happen

1663
01:00:46,080 --> 01:00:48,400
before any calculation even starts.

1664
01:00:48,400 --> 01:00:50,560
Somebody has to define the priority rules.

1665
01:00:50,560 --> 01:00:53,280
Is the plant protecting confirmed customer dates above all else?

1666
01:00:53,280 --> 01:00:56,000
Are we trying to keep a constrained resource running flat out?

1667
01:00:56,000 --> 01:00:59,520
Does a strategic customer get priority over a lower margin rush order?

1668
01:00:59,520 --> 01:01:01,120
How much overtime is acceptable?

1669
01:01:01,120 --> 01:01:02,560
And who gets to approve it?

1670
01:01:02,560 --> 01:01:04,960
Those are business policies, not mass problems.

1671
01:01:04,960 --> 01:01:08,160
APS can apply those rules consistently once they're written down.

1672
01:01:08,160 --> 01:01:11,120
It can test what happens when you move one order, change a resource

1673
01:01:11,120 --> 01:01:13,360
or reserve capacity for a high priority job

1674
01:01:13,360 --> 01:01:16,480
but it can't decide that one customer relationship matters more than another

1675
01:01:16,480 --> 01:01:19,600
unless the business gives it a rule that captures that choice.

1676
01:01:19,600 --> 01:01:21,280
And even then rules have limits.

1677
01:01:21,280 --> 01:01:23,120
Picture a planner staring at two orders

1678
01:01:23,120 --> 01:01:25,520
that both need the same remaining capacity.

1679
01:01:25,520 --> 01:01:28,400
One supports a long-term customer who has a planned shutdown coming up.

1680
01:01:28,400 --> 01:01:30,400
The other belongs to a newer customer

1681
01:01:30,400 --> 01:01:33,200
but it contains a component that will block a larger assembly

1682
01:01:33,200 --> 01:01:35,840
if it arrives late. A schedule can calculate dates,

1683
01:01:35,840 --> 01:01:37,840
cues, setup time and capacity use.

1684
01:01:37,840 --> 01:01:41,360
What it can't do is read the room after a commercial meeting

1685
01:01:41,360 --> 01:01:44,240
or understand a promise someone made during a phone call

1686
01:01:44,240 --> 01:01:47,920
unless that information somehow gets fed into the decision process.

1687
01:01:47,920 --> 01:01:49,920
That judgment belongs to the planner.

1688
01:01:49,920 --> 01:01:51,600
Now I hear people talk about this wrong sometimes.

1689
01:01:51,600 --> 01:01:55,280
They describe the planner's job as dragging work orders around a screen all day

1690
01:01:55,280 --> 01:01:59,440
and frame APS as the software that finally frees them from that primitive task.

1691
01:01:59,440 --> 01:02:00,720
There's some truth in that.

1692
01:02:00,720 --> 01:02:05,200
Manual rescheduling does eat-up time especially when one change ripples through dozens of operations.

1693
01:02:05,200 --> 01:02:07,040
But moving orders was never the whole job.

1694
01:02:07,040 --> 01:02:09,920
A good planner is constantly weighing competing needs.

1695
01:02:09,920 --> 01:02:11,440
They protect flow where it matters.

1696
01:02:11,440 --> 01:02:14,480
They ask whether a local fix just pushes trouble downstream.

1697
01:02:14,480 --> 01:02:17,440
They know when a calculated answer needs a supervisor's input,

1698
01:02:17,440 --> 01:02:21,360
a quality review or a conversation with sales before anyone changes a commitment.

1699
01:02:21,360 --> 01:02:25,120
So what APS actually does is change the speed and consistency of that work.

1700
01:02:25,120 --> 01:02:28,480
Instead of manually rebuilding every possible knock-on effect,

1701
01:02:28,480 --> 01:02:32,000
the planner can ask the system to test a defined scenario.

1702
01:02:32,000 --> 01:02:34,000
What happens if this resource stays unavailable?

1703
01:02:34,000 --> 01:02:36,080
What changes if this order gets priority?

1704
01:02:36,080 --> 01:02:39,040
Which customer dates shift if we keep the existing sequence?

1705
01:02:39,040 --> 01:02:42,320
The system runs those checks without losing track of linked operations

1706
01:02:42,320 --> 01:02:45,600
or applying a different rule halfway through because someone is under pressure.

1707
01:02:45,600 --> 01:02:46,240
That's useful.

1708
01:02:46,240 --> 01:02:48,640
It gives the planner time to make actual decisions.

1709
01:02:48,640 --> 01:02:52,720
Even then, there are cases where human review absolutely has to stay in the loop.

1710
01:02:52,720 --> 01:02:57,120
A safety concern doesn't become less serious because an optimizer found free capacity.

1711
01:02:57,120 --> 01:03:01,920
An unusual quality issue may require a hold that nobody should override through a planning rule.

1712
01:03:01,920 --> 01:03:05,120
A customer commitment can involve contract terms, commercial risk,

1713
01:03:05,120 --> 01:03:08,080
or a relationship that the scheduling model simply doesn't know about.

1714
01:03:08,080 --> 01:03:10,800
The system should surface the impact, people should decide

1715
01:03:10,800 --> 01:03:12,960
when the decision crosses those boundaries.

1716
01:03:12,960 --> 01:03:16,400
Trust also depends on whether APS can actually explain its answer.

1717
01:03:16,400 --> 01:03:20,400
If the schedule moves in order from Tuesday to Thursday, the planner needs to know why.

1718
01:03:20,400 --> 01:03:22,160
Was this selected machine unavailable?

1719
01:03:22,160 --> 01:03:24,160
Did a higher priority order take the slot?

1720
01:03:24,160 --> 01:03:25,920
Did a setup rule change the sequence?

1721
01:03:25,920 --> 01:03:27,760
Did the model find a material constrained?

1722
01:03:27,760 --> 01:03:29,920
A labor limit or a root dependency?

1723
01:03:29,920 --> 01:03:33,600
Because the algorithm decided is not going to survive a production meeting.

1724
01:03:33,600 --> 01:03:36,240
Explainability also helps expose bad assumptions.

1725
01:03:36,240 --> 01:03:39,680
When a planner sees why the system plays an order in a certain position,

1726
01:03:39,680 --> 01:03:42,160
they can challenge the rule or correct the constraint.

1727
01:03:42,160 --> 01:03:44,400
That feedback loop improves the model over time.

1728
01:03:44,400 --> 01:03:46,480
A black box recommendation does the opposite.

1729
01:03:46,480 --> 01:03:49,920
It turns the planner into someone who either accepts an answer they can't defend

1730
01:03:49,920 --> 01:03:52,720
or reverts to manual work whenever pressure rises.

1731
01:03:52,720 --> 01:03:53,840
Neither outcome helps.

1732
01:03:53,840 --> 01:03:57,200
The more useful role for APS is decision support with discipline.

1733
01:03:57,200 --> 01:04:00,960
It gives planners a consistent view of consequences under agreed rules,

1734
01:04:00,960 --> 01:04:04,480
while the planner retains control over exceptions, priorities,

1735
01:04:04,480 --> 01:04:07,680
and the point at which a scenario becomes a released production plan.

1736
01:04:07,680 --> 01:04:10,400
Over time, the planner's role shifts.

1737
01:04:10,400 --> 01:04:13,760
Less time goes into manually shuffling orders after every disruption.

1738
01:04:13,760 --> 01:04:16,160
More time goes into managing constraints,

1739
01:04:16,160 --> 01:04:18,880
checking assumptions, talking through trade-offs,

1740
01:04:18,880 --> 01:04:21,200
and improving the rules that govern the schedule.

1741
01:04:21,200 --> 01:04:23,040
That's a better use of planning expertise

1742
01:04:23,040 --> 01:04:25,760
because it turns experience into a repeatable process

1743
01:04:25,760 --> 01:04:28,240
instead of leaving it trapped in one person's memory.

1744
01:04:28,240 --> 01:04:31,280
Still, even with a fully deployed APS system in place,

1745
01:04:31,280 --> 01:04:34,480
most planning rooms keep one familiar tool close by.

1746
01:04:34,480 --> 01:04:37,840
Excel or why Excel survives every planning project?

1747
01:04:37,840 --> 01:04:40,240
Excel survives because it solves a problem in the moment.

1748
01:04:40,240 --> 01:04:42,080
A planner can open a workbook,

1749
01:04:42,080 --> 01:04:45,360
copy the orders that matter, change a date, move a job,

1750
01:04:45,360 --> 01:04:47,760
and test an idea without waiting for an interface

1751
01:04:47,760 --> 01:04:50,240
an approval flow or a system change request.

1752
01:04:50,240 --> 01:04:51,760
That speed has a place.

1753
01:04:51,760 --> 01:04:55,520
When a supplier calls with a delay, a supervisor reports an unexpected limit

1754
01:04:55,520 --> 01:04:57,920
or sales asks for a quick what if check.

1755
01:04:57,920 --> 01:05:00,000
A local workbook can become the fastest way

1756
01:05:00,000 --> 01:05:03,200
to put the situation into a form people can actually discuss.

1757
01:05:03,200 --> 01:05:05,760
It also handles missing data surprisingly well,

1758
01:05:05,760 --> 01:05:07,440
at least from the planner's perspective.

1759
01:05:07,440 --> 01:05:09,840
If the formal system doesn't hold a setup rule,

1760
01:05:09,840 --> 01:05:12,880
a labor restriction, or an informal customer priority,

1761
01:05:12,880 --> 01:05:15,360
the planner can just add a column and carry on.

1762
01:05:15,360 --> 01:05:17,840
The spreadsheet becomes a temporary model of the factory.

1763
01:05:17,840 --> 01:05:20,800
One that the official systems never captured.

1764
01:05:20,800 --> 01:05:23,280
Temporary is doing a lot of work in that sentence.

1765
01:05:23,280 --> 01:05:25,760
Some spreadsheets stay temporary for years.

1766
01:05:25,760 --> 01:05:28,160
They become the place where the actual sequence lives

1767
01:05:28,160 --> 01:05:30,800
while ERP holds the approved orders and other systems

1768
01:05:30,800 --> 01:05:32,560
hold pieces of execution data.

1769
01:05:32,560 --> 01:05:35,200
Then the planner spends the day moving information between them,

1770
01:05:35,200 --> 01:05:38,000
checking whether the workbook still matches the latest order change

1771
01:05:38,000 --> 01:05:39,120
or production result.

1772
01:05:39,120 --> 01:05:40,960
People often treat that as a tool problem.

1773
01:05:40,960 --> 01:05:41,760
It usually isn't.

1774
01:05:41,760 --> 01:05:46,560
A planner trusts a spreadsheet because they can follow the calculation.

1775
01:05:46,560 --> 01:05:49,520
They can see the rows inspect the formulas, change an assumption

1776
01:05:49,520 --> 01:05:51,280
and explain why a date moved.

1777
01:05:51,280 --> 01:05:54,320
Even when the workbook is fragile, it's logic feels visible.

1778
01:05:54,320 --> 01:05:56,000
That matters when someone from sales asks

1779
01:05:56,000 --> 01:05:58,400
why an urgent order can't run tomorrow.

1780
01:05:58,400 --> 01:06:00,960
Opac system logic creates the opposite reaction.

1781
01:06:00,960 --> 01:06:02,960
If a scheduling tool produces a new sequence

1782
01:06:02,960 --> 01:06:04,720
but doesn't show the resource conflict,

1783
01:06:04,720 --> 01:06:07,040
setup condition, or priority rule behind it,

1784
01:06:07,040 --> 01:06:09,360
the planner has no way to defend the answer.

1785
01:06:09,360 --> 01:06:12,080
Under pressure, they will return to the model they can inspect.

1786
01:06:12,080 --> 01:06:13,440
That isn't stubbornness.

1787
01:06:13,440 --> 01:06:14,400
It's accountability.

1788
01:06:14,400 --> 01:06:19,120
A planner who releases work needs to explain the consequences to production,

1789
01:06:19,440 --> 01:06:21,600
purchasing, and customer service.

1790
01:06:21,600 --> 01:06:23,680
The system chose it is not a useful answer

1791
01:06:23,680 --> 01:06:25,920
when a supervisor asks why a job with material

1792
01:06:25,920 --> 01:06:27,760
ready to run has dropped down the list.

1793
01:06:27,760 --> 01:06:30,880
So Excel remains because it gives people control over exceptions.

1794
01:06:30,880 --> 01:06:33,120
It's quick, familiar, and open enough to absorb facts

1795
01:06:33,120 --> 01:06:34,480
that haven't found a proper home.

1796
01:06:34,480 --> 01:06:37,120
But the workbook also tells you something about the process.

1797
01:06:37,120 --> 01:06:39,680
If planners repeatedly add the same extra columns,

1798
01:06:39,680 --> 01:06:41,680
those columns point to missing master data.

1799
01:06:41,680 --> 01:06:44,000
If every change needs a manual email or phone call

1800
01:06:44,000 --> 01:06:46,560
before someone can act, that points to a workflow gap.

1801
01:06:46,560 --> 01:06:48,400
If planners can change the priority list

1802
01:06:48,400 --> 01:06:51,360
but nobody knows who has authority to change a customer promise,

1803
01:06:51,360 --> 01:06:53,360
that points to unclear decision rights.

1804
01:06:53,360 --> 01:06:55,840
The spreadsheet is often the smoke, not the fire.

1805
01:06:55,840 --> 01:06:56,800
Take a common case.

1806
01:06:56,800 --> 01:06:59,040
A planner keeps a separate tab for tooling

1807
01:06:59,040 --> 01:07:02,000
because the ERP route names the machine but not the fixture.

1808
01:07:02,000 --> 01:07:04,240
Each morning, they check which fixture sits where,

1809
01:07:04,240 --> 01:07:06,240
then adjust the sequence by hand.

1810
01:07:06,240 --> 01:07:08,400
Replacing that workbook with an APS screen

1811
01:07:08,400 --> 01:07:10,720
won't solve anything unless the new planning process

1812
01:07:10,720 --> 01:07:12,720
can represent fixture availability

1813
01:07:12,720 --> 01:07:14,400
and keep that information current.

1814
01:07:14,400 --> 01:07:16,080
Otherwise, the planner will keep the tab,

1815
01:07:16,080 --> 01:07:18,000
the same applies to what if analysis.

1816
01:07:18,000 --> 01:07:20,880
A good APS process should let planners test an outage,

1817
01:07:20,880 --> 01:07:22,960
a rush order or a change material date

1818
01:07:22,960 --> 01:07:24,800
without corrupting the release plan.

1819
01:07:24,800 --> 01:07:26,640
It should show the effect of the scenario

1820
01:07:26,640 --> 01:07:28,240
in terms people can challenge,

1821
01:07:28,240 --> 01:07:30,800
which work moved, which constraint caused the delay,

1822
01:07:30,800 --> 01:07:32,080
what assumption changed.

1823
01:07:32,080 --> 01:07:34,320
Until the new process handles those hard cases,

1824
01:07:34,320 --> 01:07:36,960
removing Excel just removes the planner's safety net.

1825
01:07:36,960 --> 01:07:38,640
That doesn't mean every local workbook

1826
01:07:38,640 --> 01:07:41,200
deserves to become a formal system requirement.

1827
01:07:41,200 --> 01:07:43,040
Some are personal aids, some exist

1828
01:07:43,040 --> 01:07:45,520
because one person likes a particular view,

1829
01:07:45,520 --> 01:07:48,240
others simply duplicate a report that the organization

1830
01:07:48,240 --> 01:07:50,480
could produce more reliably elsewhere.

1831
01:07:50,480 --> 01:07:52,800
You need to separate convenience from dependency.

1832
01:07:52,800 --> 01:07:55,040
Ask planners to show you the workbook they reach for

1833
01:07:55,040 --> 01:07:56,480
when the normal plan fails.

1834
01:07:56,480 --> 01:07:58,240
Then ask what decision it supports,

1835
01:07:58,240 --> 01:08:00,560
what data it adds, where that data comes from,

1836
01:08:00,560 --> 01:08:03,200
and what happens if the workbook owner is unavailable.

1837
01:08:03,200 --> 01:08:05,040
You will learn more from that conversation

1838
01:08:05,040 --> 01:08:07,680
than from a generic list of desired APS functions.

1839
01:08:07,680 --> 01:08:09,280
And Excel has another advantage

1840
01:08:09,280 --> 01:08:10,880
that software projects sometimes forget.

1841
01:08:10,880 --> 01:08:13,440
It rarely fails in audit because it lacks a feature.

1842
01:08:13,440 --> 01:08:16,640
It fails because nobody can prove which version drove the decision

1843
01:08:16,640 --> 01:08:18,000
where its inputs came from,

1844
01:08:18,000 --> 01:08:20,080
or who changed a formula after the plan went out.

1845
01:08:20,080 --> 01:08:21,520
That's the real problem to solve.

1846
01:08:21,520 --> 01:08:24,400
A schedule people trust needs more than a better planning engine.

1847
01:08:24,400 --> 01:08:26,240
It needs clear ownership of the facts,

1848
01:08:26,240 --> 01:08:28,720
the constraints, the changes, and the release decision.

1849
01:08:28,720 --> 01:08:30,720
And from there we can start looking at the architecture

1850
01:08:30,720 --> 01:08:32,080
behind that trust.

1851
01:08:32,080 --> 01:08:35,600
Architecture, one source of record, several sources of truth.

1852
01:08:35,600 --> 01:08:38,400
Once you stop relying on Excel as your unofficial control tower,

1853
01:08:38,400 --> 01:08:40,720
the architecture needs clear ownership.

1854
01:08:40,720 --> 01:08:42,640
I'm not saying one system owns everything

1855
01:08:42,640 --> 01:08:44,240
that usually creates more confusion.

1856
01:08:44,240 --> 01:08:46,160
But one agreed owner for each type of fact

1857
01:08:46,160 --> 01:08:47,200
and each type of decision.

1858
01:08:47,200 --> 01:08:50,400
ERP stays the source of record for commercial and supply transactions.

1859
01:08:50,400 --> 01:08:53,200
It owns the customer order, the approved product data,

1860
01:08:53,200 --> 01:08:55,840
planned and firm supply, purchasing commitments,

1861
01:08:55,840 --> 01:08:57,840
inventory movements, costs,

1862
01:08:57,840 --> 01:08:59,840
and the formal status of a production order.

1863
01:08:59,840 --> 01:09:01,920
That record matters because sales needs to know

1864
01:09:01,920 --> 01:09:03,200
what the customer ordered.

1865
01:09:03,200 --> 01:09:05,920
Purchasing needs a controlled view of demand,

1866
01:09:05,920 --> 01:09:08,080
finance needs traceable transactions,

1867
01:09:08,080 --> 01:09:10,880
and the warehouse needs to know what stock can move,

1868
01:09:10,880 --> 01:09:13,920
what's blocked, and what already belongs to another order.

1869
01:09:13,920 --> 01:09:18,000
ERP doesn't need to own every operational detail to play that role.

1870
01:09:18,000 --> 01:09:20,960
MES owns execution facts from the shop floor.

1871
01:09:20,960 --> 01:09:23,040
It records what production actually did,

1872
01:09:23,040 --> 01:09:25,760
an operation starting, a quantity completing,

1873
01:09:25,760 --> 01:09:27,600
scrap material consumed,

1874
01:09:27,600 --> 01:09:29,040
or a line becoming unavailable.

1875
01:09:29,040 --> 01:09:32,240
Those facts need to come from as close to the work as possible.

1876
01:09:32,240 --> 01:09:34,480
A planner might expect an order to finish by noon,

1877
01:09:34,480 --> 01:09:36,960
but MES records that the operation still hasn't started

1878
01:09:36,960 --> 01:09:38,640
because the job is waiting for a release,

1879
01:09:38,640 --> 01:09:40,400
a setup, or a machine recovery.

1880
01:09:40,400 --> 01:09:42,080
That isn't a planning opinion.

1881
01:09:42,080 --> 01:09:43,600
It's a production fact.

1882
01:09:43,600 --> 01:09:46,240
APS owns a different kind of truth altogether.

1883
01:09:46,240 --> 01:09:47,680
It owns the constrained schedule

1884
01:09:47,680 --> 01:09:49,760
and the planning scenarios built from current rules,

1885
01:09:49,760 --> 01:09:51,760
resource availability, and order priorities.

1886
01:09:51,760 --> 01:09:54,000
It answers questions like if we reserve this machine

1887
01:09:54,000 --> 01:09:56,240
for the urgent order, which other orders move.

1888
01:09:56,240 --> 01:09:58,400
If material arrives tomorrow instead of today,

1889
01:09:58,400 --> 01:09:59,920
where does the schedule break?

1890
01:09:59,920 --> 01:10:01,840
If we add an approved overtime shift,

1891
01:10:01,840 --> 01:10:03,360
which customer dates can we protect?

1892
01:10:03,360 --> 01:10:05,040
Those answers change as conditions change.

1893
01:10:05,040 --> 01:10:06,720
They aren't permanent business transactions

1894
01:10:06,720 --> 01:10:07,840
and shouldn't pretend to be.

1895
01:10:07,840 --> 01:10:09,120
They're controlled planning views

1896
01:10:09,120 --> 01:10:10,720
that let people test the future

1897
01:10:10,720 --> 01:10:12,640
before committing the factory to a sequence.

1898
01:10:12,640 --> 01:10:14,400
That distinction feels uncomfortable at first

1899
01:10:14,400 --> 01:10:16,640
because everyone wants one date they can trust.

1900
01:10:16,640 --> 01:10:19,600
You should want one agreed release date for a given purpose,

1901
01:10:19,600 --> 01:10:21,440
but you shouldn't force every system

1902
01:10:21,440 --> 01:10:22,800
to report the same date

1903
01:10:22,800 --> 01:10:24,640
when each one describes something different.

1904
01:10:24,640 --> 01:10:26,640
ERP holds the contractual due date.

1905
01:10:26,640 --> 01:10:29,360
APS calculates the current feasible completion

1906
01:10:29,360 --> 01:10:33,120
and MES shows expected completion based on work already underway.

1907
01:10:33,120 --> 01:10:35,200
Those dates can differ without any system being wrong.

1908
01:10:35,200 --> 01:10:37,920
The problem starts when nobody knows what each date means.

1909
01:10:37,920 --> 01:10:40,720
IoT data adds another layer where it makes sense.

1910
01:10:40,720 --> 01:10:42,720
A connected asset reports a state change

1911
01:10:42,720 --> 01:10:45,680
and alarm a condition issue or loss of production rate.

1912
01:10:45,680 --> 01:10:48,560
That information gives APS or MES early warning

1913
01:10:48,560 --> 01:10:50,480
that a capacity assumption no longer holds,

1914
01:10:50,480 --> 01:10:53,440
but IoT doesn't become the owner of production truth

1915
01:10:53,440 --> 01:10:55,440
just because it generates data quickly.

1916
01:10:55,440 --> 01:10:57,120
A sensor reports that a machine stopped

1917
01:10:57,120 --> 01:10:59,200
but it doesn't know whether an operator restarted it,

1918
01:10:59,200 --> 01:11:00,880
maintenance declared it unavailable

1919
01:11:00,880 --> 01:11:03,120
or the stop effects the released schedule.

1920
01:11:03,120 --> 01:11:05,040
That event needs context and a business rule

1921
01:11:05,040 --> 01:11:06,560
before it becomes a planning action.

1922
01:11:06,560 --> 01:11:08,240
That's why clear ownership matters more

1923
01:11:08,240 --> 01:11:09,760
than a long list of integrations.

1924
01:11:09,760 --> 01:11:12,720
Picture a planner trying to answer a simple customer question.

1925
01:11:12,720 --> 01:11:14,800
Can we still ship on Friday?

1926
01:11:14,800 --> 01:11:17,680
ERP shows the original promise APS shows a constraint projection

1927
01:11:17,680 --> 01:11:19,040
after yesterday's outage.

1928
01:11:19,040 --> 01:11:21,600
MES shows the current operation began later than planned

1929
01:11:21,600 --> 01:11:24,880
and a maintenance system reports a resource restriction for tomorrow.

1930
01:11:24,880 --> 01:11:26,880
Every system contributes something useful.

1931
01:11:26,880 --> 01:11:28,400
The planner shouldn't have to reconcile

1932
01:11:28,400 --> 01:11:30,160
four competing answers by hand.

1933
01:11:30,160 --> 01:11:33,360
The architecture should define which facts feed the schedule

1934
01:11:33,360 --> 01:11:35,760
who approves a change to the customer commitment

1935
01:11:35,760 --> 01:11:38,160
and where the released production sequence lives

1936
01:11:38,160 --> 01:11:40,560
after that decision that gives people a path

1937
01:11:40,560 --> 01:11:41,840
through the disagreement.

1938
01:11:41,840 --> 01:11:44,000
Execution facts update the planning model,

1939
01:11:44,000 --> 01:11:45,680
planning exposes the impact

1940
01:11:45,680 --> 01:11:48,000
and the appropriate business owner accepts, rejects

1941
01:11:48,000 --> 01:11:49,680
or escalates the change.

1942
01:11:49,680 --> 01:11:51,680
ERP records the govern transaction

1943
01:11:51,680 --> 01:11:54,720
when that decision affects the commercial or supply record.

1944
01:11:54,720 --> 01:11:56,560
Without those boundaries, dates multiply.

1945
01:11:56,560 --> 01:12:00,080
Someone exports a list from ERP, someone else trusts the MES status

1946
01:12:00,080 --> 01:12:02,720
and a third person works from an APS scenario

1947
01:12:02,720 --> 01:12:04,160
that never became active.

1948
01:12:04,160 --> 01:12:07,280
Then the planning meeting becomes a debate about whose data is correct.

1949
01:12:07,280 --> 01:12:10,640
A strangely expensive way to avoid deciding what a date represents.

1950
01:12:10,640 --> 01:12:12,800
Clear ownership doesn't remove disagreement.

1951
01:12:12,800 --> 01:12:15,600
It makes disagreement visible in the right place with the right people

1952
01:12:15,600 --> 01:12:18,000
before production discovers it through missed work.

1953
01:12:18,000 --> 01:12:19,440
Once those system roles are clear,

1954
01:12:19,440 --> 01:12:21,040
another assumption needs attention.

1955
01:12:21,040 --> 01:12:24,880
Connecting ERP, APS, MES and IoT data

1956
01:12:24,880 --> 01:12:26,880
doesn't automatically give you a planning model

1957
01:12:26,880 --> 01:12:28,480
that understands the factory.

1958
01:12:28,480 --> 01:12:31,440
Data integration does not create manufacturing context.

1959
01:12:31,440 --> 01:12:34,560
You can connect ERP, APS, MES and machine data

1960
01:12:34,560 --> 01:12:36,160
into one shared data platform

1961
01:12:36,160 --> 01:12:38,000
and still fail to produce a schedule that works.

1962
01:12:38,000 --> 01:12:40,240
That sounds odd at first because integration work

1963
01:12:40,240 --> 01:12:41,840
often feels like the hard part.

1964
01:12:41,840 --> 01:12:44,320
You build interfaces, map fields, handle errors,

1965
01:12:44,320 --> 01:12:45,440
match order numbers,

1966
01:12:45,440 --> 01:12:47,680
and finally bring the data into one place.

1967
01:12:47,680 --> 01:12:49,920
Now you can query an order, see its routing,

1968
01:12:49,920 --> 01:12:51,120
find the machine's status,

1969
01:12:51,120 --> 01:12:52,800
and check the material position.

1970
01:12:52,800 --> 01:12:55,840
But connected records don't automatically explain their relationships.

1971
01:12:55,840 --> 01:12:57,680
A production order links to a routing,

1972
01:12:57,680 --> 01:12:59,440
that routing lists an operation

1973
01:12:59,440 --> 01:13:01,840
and that operation points to a work center.

1974
01:13:01,840 --> 01:13:04,160
Yet a planner needs a more specific answer,

1975
01:13:04,160 --> 01:13:06,880
which exact machine can perform this operation

1976
01:13:06,880 --> 01:13:08,800
under what conditions with which tool,

1977
01:13:08,800 --> 01:13:10,320
using which qualified worker,

1978
01:13:10,320 --> 01:13:12,640
and only after which material becomes usable,

1979
01:13:12,640 --> 01:13:15,120
that is manufacturing context.

1980
01:13:15,120 --> 01:13:17,200
Consider a typical operation in a plant.

1981
01:13:17,200 --> 01:13:20,480
The ERP record tells you that order 1842 needs milling.

1982
01:13:20,480 --> 01:13:22,480
MES shows the order hasn't started,

1983
01:13:22,480 --> 01:13:23,760
a machine data source reports

1984
01:13:23,760 --> 01:13:26,480
that two milling machines are running and one is idle.

1985
01:13:26,480 --> 01:13:28,480
None of those facts tell you whether the idle machine

1986
01:13:28,480 --> 01:13:30,240
can run order 1842.

1987
01:13:30,240 --> 01:13:32,240
Maybe it lacks the required fixture,

1988
01:13:32,240 --> 01:13:35,040
or the operator on shift isn't qualified for that product.

1989
01:13:35,040 --> 01:13:36,880
Maybe a tool is installed for another job

1990
01:13:36,880 --> 01:13:39,600
and changing it would delay an order with a tighter due date.

1991
01:13:39,600 --> 01:13:41,680
Or maybe the machine can process the part,

1992
01:13:41,680 --> 01:13:44,720
but only after a heat treatment step releases the material.

1993
01:13:44,720 --> 01:13:47,040
The data exists, the relationship is missing.

1994
01:13:47,040 --> 01:13:49,040
A context model puts those relationships

1995
01:13:49,040 --> 01:13:51,280
into a form the planning process can use.

1996
01:13:51,280 --> 01:13:53,520
It describes resources, operations, tools,

1997
01:13:53,520 --> 01:13:55,840
labor, materials, and the connecting rules.

1998
01:13:55,840 --> 01:13:57,760
More importantly, it describes limits.

1999
01:13:57,760 --> 01:14:00,160
A resource isn't simply available or unavailable,

2000
01:14:00,160 --> 01:14:01,840
it may be available for some work,

2001
01:14:01,840 --> 01:14:03,520
during some shifts with some people,

2002
01:14:03,520 --> 01:14:05,360
and under some sequence conditions.

2003
01:14:05,360 --> 01:14:08,000
That is what makes a schedule more than a list of dates.

2004
01:14:08,000 --> 01:14:11,680
A practical way to structure this is through a product process resource model.

2005
01:14:11,680 --> 01:14:12,880
Product means what you're building,

2006
01:14:12,880 --> 01:14:15,680
including the material and product rules that matter to production.

2007
01:14:15,680 --> 01:14:18,640
Process means the sequence of operations needed to build it.

2008
01:14:18,640 --> 01:14:20,720
Resource means what performs those operations.

2009
01:14:20,720 --> 01:14:23,280
Machines, lines, tools, fixtures, people,

2010
01:14:23,280 --> 01:14:26,720
and sometimes shared services like inspection or heat treatment.

2011
01:14:26,720 --> 01:14:28,720
The model connects those three areas.

2012
01:14:28,720 --> 01:14:31,040
For a given product, which process pathways?

2013
01:14:31,040 --> 01:14:33,760
For each operation, which resources can do the work?

2014
01:14:33,760 --> 01:14:35,920
What capacity does each resource have?

2015
01:14:35,920 --> 01:14:39,760
Which material, tool, qualification, or setup rule limits the choice?

2016
01:14:39,760 --> 01:14:40,880
Once those links exist,

2017
01:14:40,880 --> 01:14:43,040
APS can test a plan against the conditions

2018
01:14:43,040 --> 01:14:44,400
that shape real production.

2019
01:14:44,400 --> 01:14:47,600
Without those links, the system falls back to broad assumptions.

2020
01:14:47,600 --> 01:14:50,080
It might know that a work center has capacity

2021
01:14:50,080 --> 01:14:52,080
while missing that the only usable machine

2022
01:14:52,080 --> 01:14:53,840
already runs a long campaign.

2023
01:14:53,840 --> 01:14:56,320
It might know that material sits in inventory

2024
01:14:56,320 --> 01:14:58,560
while missing that the specific lot remains blocked

2025
01:14:58,560 --> 01:15:00,480
or hasn't reached the point of use.

2026
01:15:00,480 --> 01:15:02,720
It might assign a job to an alternate route

2027
01:15:02,720 --> 01:15:04,240
because the routing allows it,

2028
01:15:04,240 --> 01:15:08,240
while missing the labor rule that makes that alternate route unavailable this week.

2029
01:15:08,240 --> 01:15:10,160
That isn't a failure of scheduling math.

2030
01:15:10,160 --> 01:15:12,560
The schedule can only reason from the context you give it.

2031
01:15:12,560 --> 01:15:16,080
If the factory model treats all machines in a group as interchangeable,

2032
01:15:16,080 --> 01:15:18,720
APS will plan as if they are interchangeable.

2033
01:15:18,720 --> 01:15:20,800
If the model ignores a shared fixture,

2034
01:15:20,800 --> 01:15:23,520
that fixture appears available everywhere at once.

2035
01:15:23,520 --> 01:15:26,000
An impressive feat, only if you don't own the fixture.

2036
01:15:26,960 --> 01:15:29,920
This is where digital twin and knowledge graph language can help,

2037
01:15:29,920 --> 01:15:31,360
provided we keep it grounded.

2038
01:15:31,360 --> 01:15:34,720
A digital twin in this setting means a usable model of the factory objects

2039
01:15:34,720 --> 01:15:36,080
and their current state.

2040
01:15:36,080 --> 01:15:40,080
A knowledge graph means a way to model and query the links between those objects.

2041
01:15:40,080 --> 01:15:42,560
Neither phrase fixes a planning problem on its own.

2042
01:15:42,560 --> 01:15:44,800
They describe an approach,

2043
01:15:44,800 --> 01:15:46,400
represent the things in the factory

2044
01:15:46,400 --> 01:15:49,040
and the relationships that determine what can happen next.

2045
01:15:49,040 --> 01:15:51,520
You don't need to brand the project as a digital twin program

2046
01:15:51,520 --> 01:15:52,960
to gain from that approach.

2047
01:15:52,960 --> 01:15:55,920
If your APS model knows that a product needs a process,

2048
01:15:55,920 --> 01:15:57,920
that process needs a qualified resource

2049
01:15:57,920 --> 01:16:00,640
and that resource depends on a tool and available material,

2050
01:16:00,640 --> 01:16:04,480
you're already modeling the relationships that a twin or graph would make explicit.

2051
01:16:04,480 --> 01:16:07,600
The name matters far less than whether the model supports the decision.

2052
01:16:07,600 --> 01:16:11,280
APS relies on this context even when nobody mentions product process resource,

2053
01:16:11,280 --> 01:16:14,080
digital twin, or knowledge graph in the project meetings.

2054
01:16:14,080 --> 01:16:17,760
Constraint based scheduling depends on knowing what constraints the work.

2055
01:16:17,760 --> 01:16:20,160
Otherwise, it simply turns disconnected records

2056
01:16:20,160 --> 01:16:22,480
into a more elaborate version of wishful planning.

2057
01:16:22,480 --> 01:16:24,480
Now let's connect that architecture to Microsoft,

2058
01:16:24,480 --> 01:16:25,840
but keep the roles clear.

2059
01:16:25,840 --> 01:16:28,080
Microsoft can help carry, govern, analyze,

2060
01:16:28,080 --> 01:16:30,000
and act on the information around the schedule.

2061
01:16:30,000 --> 01:16:31,840
It doesn't remove the need for the industrial model

2062
01:16:31,840 --> 01:16:34,240
that tells the system what a factory can actually do.

2063
01:16:34,240 --> 01:16:37,280
Where Microsoft fits in an APS and ERP architecture,

2064
01:16:37,280 --> 01:16:39,680
Microsoft sits around the planning process,

2065
01:16:39,680 --> 01:16:41,360
not inside the planning logic itself.

2066
01:16:41,360 --> 01:16:42,720
That sounds like a small difference,

2067
01:16:42,720 --> 01:16:45,200
but it heads off a lot of architecture confusion

2068
01:16:45,200 --> 01:16:47,600
when you're trying to connect enterprise data,

2069
01:16:47,600 --> 01:16:50,240
shop floor events, and a constrained production schedule.

2070
01:16:50,960 --> 01:16:53,520
Azure can act as the integration layer between systems

2071
01:16:53,520 --> 01:16:55,200
that own different parts of the process.

2072
01:16:55,200 --> 01:16:57,440
It can move approved order changes out of ERP,

2073
01:16:57,440 --> 01:17:00,320
receive production events from MS or connected equipment,

2074
01:17:00,320 --> 01:17:02,720
and root the right information to the APS layer

2075
01:17:02,720 --> 01:17:05,760
without giving every system direct access to everything else.

2076
01:17:05,760 --> 01:17:08,160
That matters for both security and control.

2077
01:17:08,160 --> 01:17:10,560
A production event should carry enough information for the next few days.