TOKYO — At a General Motors plant outside Detroit, robots affixing windshields to cars or painting chassis call out across the floor, suggesting ways to do their jobs more efficiently. They are practicing kaizen, or continuous incremental improvement, with no need for cues from human minders.
This is the American automaker’s vision of the factory of the future — a goal it aims to achieve within five years. Artificial intelligence-equipped robots armed with data from GM’s global production network come up with ways to build vehicles faster and more cheaply. Productivity should grow at an accelerating pace, massively boosting GM’s competitive edge, said Marty Linn, manager of advanced technology.
The painful experience of being forced to shutter plants en masse after filing for Chapter 11 bankruptcy protection in 2009 has driven GM to work toward realizing this dream. The automaker, in partnership with Cisco Systems and Japanese industrial-robot maker Fanuc, developed a system in which more than 8,500 robots worldwide share information every 90 seconds. Last year, 65 of those robots warned that they would fail in two weeks, letting GM head off the problem.
A century ago, the American auto industry introduced mass production with workers aided by machines, slashing prices and making cars widely accessible to the public. Now AI is poised to bring about a new revolution in manufacturing.
Beer by algorithm
The beer gave off a pleasant apple aroma but left a slightly bitter aftertaste.
Version 13 would have yet another different flavor, said Intelligentx Brewing co-founder Hew Leith, explaining that the London brewery’s AI would take into account the drinker’s feedback when developing the next recipe. The beer has gone through 12 iterations over the last year and a half, with the AI refining the formula each time.
AI is poised to change even how beer lovers shop, eliminating the need to choose from the limited options on store shelves. Intelligentx’s AI solicits impressions and requests from drinkers via Facebook and uses this input to calculate the appropriate flavor and feel for the next version. It then comes up with the new recipe, adjusting such factors as hops content and foam formation.
It will become possible to develop beverages tailored to each customer, Leith said. Restaurants and individuals will be able to quickly develop and offer their own varieties. The age of customized beer is nearly upon us.