Introduction: Toyota in 1985

In 1985 General Motors opened a plant in Hamtramck, Detroit, that was meant to settle the argument about technology and competitiveness for good. It was billed as the factory of the future. It held some 260 robots for welding, painting and assembly, automated guided vehicles to ferry parts, and cameras and lasers to inspect the work. Roger Smith, the chairman who commissioned it, was in the middle of a spending programme on automation and acquisitions that would run to roughly 45 billion dollars over the decade, a sum large enough at the time to have simply bought Toyota instead.

The plant was a disaster. The robots smashed windscreens, painted each other, and welded doors shut. Software crashed the body shop for hours at a stretch. Workers stood and watched machines do badly what they had once done well. Hamtramck limped for years below its rated capacity, producing cars of unremarkable quality at extraordinary cost.

That same year, about 2,000 miles west in Fremont, California, Toyota was running an experiment from the opposite direction, in a plant General Motors had closed as unmanageable.

When researchers from MIT's International Motor Vehicle Program surveyed the industry in 1986, the gap was not subtle. GM's Framingham plant took 31 assembly hours to build a car. Toyota's Takaoka plant took 16. Framingham produced 135 defects per 100 cars, Takaoka 45. Framingham held two weeks of inventory, Takaoka two hours.

The lesson took a decade to sink in and a generation to act on, and it is the lesson this book is built around. GM bought technology and expected transformation. Toyota built a production system and got one.