A production system is the set of rules a factory uses to decide what to make, in what order, at what pace, with how much inventory, and what to do when something goes wrong. It is the least visible and most valuable asset a manufacturer has: it cannot be bought with the machines, and it is the part of process capability that competitors find hardest to copy. Ford's moving assembly line at Highland Park (1913, from memory, verify) made the car cheap by making one product the same way at enormous volume. General Motors under Alfred Sloan added variety and management by numbers. Toyota, too small and short of capital after the war to copy Ford, built a system under Taiichi Ohno from the late 1940s through the 1970s that made small batches of many models with little inventory and stopped the line at every defect: just-in-time, kanban, jidoka, levelled production and continuous improvement. The clearest proof that the system, not the workforce or the machines, made the difference came from Fremont, California. GM had closed its plant there in 1982; NUMMI, the GM–Toyota joint venture agreed in February 1983, reopened it with largely the same UAW workforce and began production in December 1984, and within a few years reached quality and productivity far above the old plant's (MIT research, reported; equal to Supply chain: Toyota). Others turned the methods into management systems: Motorola's and GE's Six Sigma, the Danaher Business System, Illinois Tool Works' 80/20 — whose "enterprise initiatives" the company credited with 130 basis points of its 26.3% operating margin in 2025 (company) — and GE Aerospace's "FLIGHT DECK". The lesson runs the other way too. Boeing's 2024 door-plug accident and the FAA's production cap showed what happens when a production system gives way to schedule pressure (03). This section explains the main production systems, what they are for, how they spread and failed to spread, what lean is not, and how to judge a company's claims about its "operating system". The companion worked example computes takt time, line balancing and work-in-process (Little's law) for one stylized assembly cell.
What the study covers
- 1Overview: the system is the product
- 2Mass production: Ford and Sloan
- 3The Toyota Production System
- 4NUMMI and the transfer problem
- 5Deming, Juran and the quality movement
- 6Six Sigma
- 7Lean outside Japan: business systems
- 8Flow arithmetic: takt, bottlenecks and Little's law
- 9What lean is not
- 10How to judge a company's "operating system"
- 11Key numbers
- 12Landmark companies and cases
- 13Pitfalls
- 14How an expert thinks about it
- 15Glossary
Technical exhibits
Worked examples, checklists and case studies included with this study.
Takt Time Line Balancing and Littles Law Worked Example
A stylized 13-task pump-assembly cell: takt time at four demand levels, the minimum and achievable number of stations, balance efficiency, labour cost per unit, work-in-process and lead time under one-piece flow and batches (Little's law), and how queues explode at the bottleneck as utilisation approaches 100% (computed in Python)
Companies in this study
- Toyota Production System
- NUMMI
- Ford Motor
- General Motors
- Tesla Manufacturing
- Boeing
- General Electric
- Danaher
- Illinois Tool Works
- Honeywell
- Parker Hannifin
- Eaton
- Alcoa
- Nucor
- Intel Foundry
- United Auto Workers
- Volkswagen Group
- Robert Bosch
- Airbus
- Komatsu
- Haier. Toyota's supply-chain side (kanban with suppliers, the 1997 Aisin fire, the 2011 tsunami and its strategic stocks) is told in Supply chain: Toyota and its 1950 move; just-in-time against just-in-case is in Supply chain: 11 Inventory Planning. The main book's Appendix A computes OEE and the cost of downtime for one line; this section's worked example is about flow and balance, not OEE.