NUMMI shows Andon-driven turnaround

A social-media thread this week revived a familiar manufacturing case: NUMMI, the GM-Toyota joint venture that reopened GM’s shuttered Fremont, California, plant in 1984. The thread argued that NUMMI’s gains in quality and attendance came less from copying tools than from changing how work was managed. Published accounts by former Toyota manager John Shook support the core point. They describe a system in which team members were expected to surface problems, team leaders were expected to respond immediately, and line stops were treated as a way to protect quality rather than a failure of discipline. ### What changed at Fremont when NUMMI opened? Toyota USA’s company history says Toyota and GM formed NUMMI in Fremont as a U.S. joint venture, and PBS reported the plant opened in 1984 after GM had shut the site in 1982. Shook wrote that NUMMI rehired many of the same workers from GM’s old Fremont operation, making the plant a test of whether a different production and management system could change outcomes with much of the same workforce. (lean.org) John Shook’s account says the shift was not framed as replacing people so much as replacing assumptions about work. In his telling, Toyota’s system defined problems as expected events to be exposed and addressed, not hidden to keep the line moving. That distinction sits at the center of most serious NUMMI retellings. (pressroom.toyota.com) ### Why does the andon cord keep showing up in NUMMI stories? The Lean Enterprise Institute’s NUMMI material describes the andon system as the clearest expression of Toyota’s approach to built-in quality. Shook wrote that the stop-the-line system signaled a basic commitment: workers were not supposed to pass defects forward, and management had to provide a rapid response when normal work broke down. (lean.org) A later Lean Enterprise Institute essay describing the “4S help chain” gave the operating detail often missing from simplified NUMMI stories. If a worker fell behind or encountered a problem, the worker pulled the andon cord, a team leader covering a small zone of stations came to help, and the issue was either resolved within the cycle or the line stopped. The point was not symbolic empowerment. It was a standard escalation path tied to a specific response. (lean.org) ### Was this about culture, or about routines? MIT Sloan Management Review’s NUMMI summary says the central cultural change in lean transformations concerns how organizations treat problems. Shook’s published NUMMI accounts make the same argument in more operational terms: behavior changed because the system of work changed first. Workers were given defined jobs, visible support, and permission—indeed an obligation—to call out abnormalities. (lean.org) Harvard Business Review’s 1993 account of Fremont said quality and productivity improved when workers defined their own job standards. That line matches the larger NUMMI record: kaizen was not an add-on suggestion program but a daily mechanism for revising work, training to the revised standard, and repeating the cycle. ### What does NUMMI say about copying Toyota systems elsewhere? (store.hbr.org) Harvard Business School’s case description says NUMMI gave GM a chance to learn the Toyota Production System up close. Yet HBR’s later reflection on NUMMI framed the venture as a case in what Toyota learned and GM did not, and Shook’s own writing has long argued that tools alone do not transfer well without the management system around them. (hbr.org) The practical lesson from the public NUMMI record is narrower than many viral threads suggest. Standard work, one-point training, escalation rules and frontline problem ownership mattered because they were linked. An andon cord without a team leader response, or kaizen without authority to change work safely, would not describe the system Shook documented at NUMMI. (store.hbr.org) ### Why does the case still get cited decades later? NUMMI closed in 2010, with PBS reporting the end of a 26-year experiment in California auto manufacturing. But the plant remains a reference point because it offered a rare controlled comparison: many of the same people, in the same place, under a different operating system. That is why current debates about lean, autonomy and stop-the-line authority keep returning to Fremont. (lean.org) John Shook’s NUMMI essays remain the most cited public source for those mechanics, and Lean Enterprise Institute still hosts several of them. For readers trying to separate folklore from operating detail, those accounts are the next stop. (lean.org) (pbs.org)