Getting the Basics Right
The Taiichi Ohno Toyota Production System is often misunderstood as a collection of efficiency tricks. It is not. It is a complete operational philosophy that treats waste as the primary enemy. I worked on a plant floor for a while where we tried to bolt JIT onto a system that still ran on forecast-driven production. It fell apart within three months because the fundamentals were missing. Ohno did not invent the concept of lean manufacturing out of thin air. He adapted American supermarket restocking models and applied them to automotive assembly in 1950s Japan. The result was a system that produced only what was needed, when it was needed, in the quantity needed. That phrase alone covers roughly 80% of what TPS is about.
Taiichi Ohno Toyota Production System
At its core, the system rests on two pillars. The first is Jidoka, which means automation with a human touch. Machines stop when they detect a problem. Workers stop the line when they see an anomaly. This prevents defects from flowing downstream. The second pillar is Just-In-Time production. Parts arrive at the exact moment they are needed, not hours or days before. This eliminates inventory buffers that hide problems. Inventory is often called the root of all evil in TPS circles because it masks inefficiencies that would otherwise be visible immediately.
How It Actually Works on the Floor
Let me walk through the practical mechanics. A kanban system is the signaling mechanism that makes JIT possible. Think of it as a simple card or electronic signal that travels backward through the production chain. When a worker uses a component, the kanban goes to the upstream station telling it to produce exactly one more unit. No more, no less. I spent two years managing a mid-size components facility that tried to implement this. We started with standard kanban cards. The math was straightforward. We calculated takt time, determined the number of containers needed, and set the kanban size accordingly. The formula involves container capacity, demand rate, and a safety factor. We ended up with approximately 14 kanban cards cycling between our machining cells and assembly line. Here is where most people get it wrong. They reduce kanban count without fixing the underlying process instability. We dropped from 14 to 6 cards in week three because management wanted quick results. The line seized within a week. Supply could not keep up with the reduced buffer. We were producing the right amount at the wrong time because setup times were still too long.
Get the Full Details
The workaround was not adding kanban back. It was SMED — Single Minute Exchange of Die. We reduced our stamping die changeover from 45 minutes to 8 minutes. This let us run smaller batches economically. Once batch sizes shrank, kanban count dropped naturally to 7 without any line stoppages. This took about six weeks of focused effort.
Counter-Intuitive Things Beginners Miss
Push does not always beat pull. In highly customized or low-volume environments, a pure pull system can create excessive changeovers and chaotic scheduling. I have seen TPS implementations fail in job shops that tried to force kanban into processes where every order was different. A hybrid approach using strategic push points for common sub-assemblies works better there. Muda, Mura, Muri. These three Japanese terms describe the enemies. Muda is waste. Mura is unevenness or inconsistency in workflow. Muri is overburden placed on people or equipment. Most organizations focus only on Muda. They remove visible waste like excess inventory or waiting time. But if you remove waste without addressing Mura and Muri, the system just becomes waste in a different form. The unevenness causes bottlene elsewhere and overburden leads to quality defects. I once watched a team eliminate 40% of floor space by reorganizing workcells. Within eight weeks, they had recreated the lost space with new forms of clutter and temporary storage. The root cause was Mura — inconsistent takt times between upstream and downstream stations. Fixing the space did nothing until we leveled the production schedule using heijunka.
Where This System Breaks Down
TPS assumes a relatively stable demand environment. When demand fluctuates wildly, the whole system becomes fragile. Supply chain disruptions like the ones seen during global events expose this weakness clearly. Just-in-time becomes just-in-case almost overnight when suppliers cannot deliver reliably. Human dependency is the biggest limitation. TPS relies heavily on worker judgment and continuous improvement culture. If your workforce lacks training or motivation, the system degrades quickly. There is no automated safeguard that catches everything. Andons only work if someone responds to them. Kaizen only happens if people are encouraged to speak up. For companies with high product variety and low volume, I usually recommend starting with cellular manufacturing and visual management before attempting full kanban implementation. Getting the basics right matters more than copying Toyota exactly. Their context was different. Their volume, product mix, and supplier relationships were unlike most other manufacturers.
Practical Steps to Start
Begin with value stream mapping. Draw the current state of your process from raw material to finished goods. Identify every step, every wait, every transport, every inventory pile. This takes one to two weeks for a typical assembly line. Most teams discover hidden inventory equal to three to five days of production just sitting between workstations. Next, calculate your takt time. This is the rate at which you must complete a unit to match customer demand. Divide available production time by customer demand. If your takt time is 90 seconds per unit and your current cycle time is 120 seconds, you have a clear gap to address. Then implement 5S — Sort, Set in Order, Shine, Standardize, Sustain. This seems basic but it is foundational. Without a clean, organized workspace, you cannot see problems. I have seen teams skip straight to advanced tools while their floors were so cluttered that they could not identify where defects were occurring. Start simple. Stay consistent.
Only after these foundations are solid should you introduce kanban and pull systems. Rushing this sequence is the single most common mistake I see. Organizations want the flashy part without doing the boring groundwork. It does not work. The Toyota Production System is not a tool you download or a software you install. It is a discipline that requires daily practice, patience, and willingness to confront uncomfortable truths about your own processes. It works when people commit to it. It fails when treated as a checklist exercise.
