Shop Floor Optimization: 12 Proven Ways to Improve Flow

Disclosure: M365.FM host Mirko Peters works at iFAKT GmbH. This page describes the company factually; paid partnerships and links are labelled as advertising.

By Mirko Peters · 2026-11-11

Shop floor optimization starts with flow. Twelve lean methods for plant managers, from takt time and bottlenecks to SMED, kanban and 5S, with a worked example.

Shop floor optimization means improving how material, information and people move through manufacturing operations, so that products leave faster and with less waiting. Better manufacturing flow raises productivity, shortens lead time and cuts stock. This guide is for plant managers who want a practical list. It gives twelve established lean methods grouped in three steps: see the flow, smooth the flow and keep it stable. A worked example with made-up numbers then shows how to choose between two ideas for one bottleneck. The sources I checked define and describe these methods, but none ranks them, so test each one on your own line.

Key points

• Shop floor optimization works best when you improve flow first: see it, smooth it and keep it stable. Faster machines come last.
• Compare each cycle time with takt time, find the bottleneck and improve that step before any other.
• Shorter changeovers, smaller batches, pull signals and levelled schedules cut waiting and stock between steps.
• 5S, standard work, visual signals and preventive maintenance keep the improved flow from sliding back.
• These are established lean methods, but the sources do not rank them. Test one change at a time and measure it.

What is shop floor optimization, and where should you start?

On the shop floor, products spend most of their time waiting between steps, not being worked on. The Lean Enterprise Institute (LEI) lexicon says that processing time is typically a small fraction of production lead time. That is why many lean methods target flow: the movement of work from step to step without stops. Improving one machine in isolation often leaves the waiting untouched. Good shop floor optimization therefore does not mean pushing people or machines harder. It means removing inefficiency in the workflow, so that you optimize the production process and enhance productivity without extra stress. Over time, the goal is a culture of continuous improvement, in which small improvements by the team are normal.

Start with a clear picture. Choose one product family and one question, such as "where does the work wait?" Then use the methods below in order. Each way has a short description, and you can find the full definitions in the sources named in the text.

12 ways to improve flow on the shop floor

Step 1: see the flow

• Way 1: Map the flow. Value stream mapping draws the material and information flow of one product family, with cycle times and stock between steps. It shows where products wait. Walk the floor to draw it.
• Way 2: Calculate takt time. Takt time is the production pace that matches customer demand: planned production time divided by customer demand, as the Lean Production glossary defines it. It is a calculation, not a measurement. Compare it with the cycle time of each step.
• Way 3: Find the bottleneck. Bottleneck analysis, in the same glossary, means identifying the part of the process that limits overall throughput and improving that part. Improving any other step does not raise output.

Step 2: smooth the flow

• Way 4: Shorten changeovers with SMED. Single-minute exchange of die is, in the words of the LEI lexicon, a process for changing production equipment from one part number to another in as little time as possible. It comes from Shigeo Shingo. The method has two steps: separate the tasks that need the machine stopped from those that can happen while it runs, then convert as many tasks as possible into the second group. The target is a single-digit number of minutes. Read the LEI entry on SMED.
• Way 5: Move towards smaller batches and one-piece flow. The Pawley Lean Institute at Oakland University, in its lean definitions, describes one-piece flow as moving work continuously one item at a time instead of in batches, which reduces waste, especially excess inventory. Full one-piece flow is not always possible, but smaller batches always help.
• Way 6: Add pull with kanban. Kanban uses signal cards that trigger restocking when more material is needed, as the Lean Production glossary describes it. Limiting the stock between steps limits the work in process.
• Way 7: Level the schedule with heijunka. Heijunka is level scheduling. The glossary says it runs small batches by mixing product variants on the same process. The Oakland definitions call it production smoothing, or levelling out production volume.
• Way 8: Release work in equal time increments. For make-to-order work, Mike Rother suggests releasing work to the chain in equal time increments based on the bottleneck's capacity, and keeping short first in, first out (FIFO) queues between steps. See his LEI article on make-to-order mapping.

Step 3: keep the flow stable

• Way 9: Make problems visible. The Lean Production glossary describes andon as a visual feedback system on the plant floor that shows production status, signals when help is needed and lets operators stop the process. Pair it with a short daily meeting at a visual board, where teams react to deviations the same day. This routine is the core of a shop floor management system, together with performance indicators and structured follow-up of problems.
• Way 10: Use 5S. 5S is a five-step method for organizing a work area: sort, straighten, shine, standardize and sustain. The first step removes items you do not need. A tidy station saves search time, helps eliminate waste and protects product quality, because tools and parts are where they belong.
• Way 11: Write standard work. The Oakland definitions describe standard work as organizing all tasks in the best-known sequence, with the best mix of people, materials, methods and machines. The glossary adds that it should be a living document that is easy to change.
• Way 12: Cut downtime with TPM and OEE. Total productive maintenance is a proactive approach that maximizes equipment uptime and involves operators in maintaining their own equipment. This cuts machine downtime. Some plants add predictive maintenance, which uses equipment data to plan repairs before a failure. Overall equipment effectiveness (OEE) combines availability, performance and quality into one of the key performance indicators for equipment. Show it on a dashboard so that the team sees downtime and utilization in real time, not in next week's report. A real-time view helps the team to streamline the response to every stop.

How do you choose where to start?

Use takt time and the bottleneck to choose, so that you optimize processes where it counts and raise productivity at the step that limits output. A method helps only if it improves the step that limits output, or the waiting in front of it. Follow this order:

1. Compare cycle times with takt time. Which steps are above takt time, or close to it?
2. Find where stock builds up. Stock piles up in front of the slowest step.
3. Ask what costs the bottleneck time. Changeovers, stops, waiting for material and rework are common losses. Choose the method that targets the biggest one.
4. Test on one line for a few weeks. Compare output, stock and lead time before and after, and track them as simple KPIs.
5. Standardize what works, then look again. After a change, the bottleneck can move to another step, so recalculate and repeat. Train the team on each new routine, because employee training keeps the gain.

Example: two ideas for one bottleneck

This is an illustrative example with made-up numbers, not a measured case. A line has a welding station as its bottleneck, with a cycle time of 85 seconds. Demand is 280 units per day. A working day has 420 net minutes. The welding station runs three product changeovers per day, and each takes 40 minutes.

Today. Changeovers use 3 × 40 = 120 minutes. The station has 420 − 120 = 300 minutes for production, or 18,000 seconds. At 85 seconds per unit, it makes 18,000 ÷ 85 = about 211 units per day. Demand is 280, so the shortfall is about 69 units.

Idea A: a faster cycle. The team trims 5 seconds from the cycle, to 80 seconds. Output is 18,000 ÷ 80 = 225 units. That is 14 more than today. Demand is still not met.

Idea B: SMED. The team applies the two SMED steps and cuts each changeover from 40 to 10 minutes. Changeovers now use 3 × 10 = 30 minutes. Production time is 420 − 30 = 390 minutes, or 23,400 seconds. At 85 seconds per unit, output is 23,400 ÷ 85 = about 275 units. That is 64 more than today, about 30 percent, and close to the demand of 280.

The comparison shows why you start with the bottleneck's lost time. Idea B recovers 90 minutes of production time. Idea A recovers a few seconds per unit. Both are worth doing. Idea B comes first, and the team measures the result before taking the next step.

Common mistakes

• Improving a step that is not the bottleneck. The extra output piles up in front of the next constraint.
• Starting with equipment. Faster machines raise cycle speed, but they do not remove waiting.
• Rolling out everything at once. Change one thing, measure it and then move on.
• Skipping the people. Operators know where the work stops. Involve them in each change.
• Treating 5S or kanban as a one-off. Without routines and follow-up, the old state returns and the productivity gain fades.
• Trusting a method because it is popular. The sources describe the methods but do not rank them. Test each one against your own data.

How software can help

Tools help when the plan changes faster than a spreadsheet can follow. iFAKT's Polaris is production software from iFAKT GmbH that combines ERP data with a digital twin of manufacturing and AI to optimize production schedules, reveal bottlenecks and improve on-time delivery. iFAKT's Polaris Optimizer is a module that automatically generates optimized production plans under constraints such as resources, shifts, zones, qualifications and capacities. A tool does not replace a walk on the floor.

Next step

Pick one product family, calculate its takt time and find its bottleneck this week. Then choose the one way from this list that targets that step, and start your shop floor optimization with a single measured test.