It was one of (too many?) Many Elon Musk’s dreams: a factory without humans. While the attention of the fans and especially financial markets is turned on Tesla, which must reveal this Wednesday, May 2 after the closure of the New York Stock Exchange (around 22 hours in Paris) its results in the first quarter of 2018, the billionaire entrepreneur has retracted his remarks.
“The human is undervalued,” he tweeted on April 13th. Even going so far as to call the overautomatization of the Fremont plant in California “error”. True “mea culpa” or simple anticipation of results that may not be up to the mark? On April 3, the brand had already announced having narrowly missed its production target for Model 3.
More humans
Without totally burying his dream, Elon Musk recognizes, however, that “the optimal level of automation remains a complex balance between design, productivity, quality and human and mechanical skills”. In an interview with CBS, on April 13, Elon Musk even announced that he planned to hire several hundred more people.
They will be added to the 10,000 employees at the Fremont plant, to reach the production level of 5,000 Model 3 per week, by the end of June. Because the entire company Tesla depends on the success of this car supposed to democratize electric vehicles to the general public.
Difficult final assembly
According to Max Warburton, an analyst at Bernstein who reported on the automation of Tesla, the problem of the firm Fremont “comes from the complexity of the final assembly” of the car. Quoted by Quartz, he explains that other major manufacturers (Fiat, Volkswagen, or General Motors) have experienced the same difficulties.
“In the final assembly, the robots […] do not detect and ignore threads that are not straight, poorly adjusted bolts, fasteners that are not aligned, or faulty seals,” he says. there. Humans are really good at this. An employee can identify what is wrong, stop the process and try to fix it.
Above all, says Bernstein’s report, “beyond a certain point, automation can increase the costs” of production. Instead of improving product quality or productivity. Maybe artificial intelligence, which should allow machines to learn from their mistakes, can bring at Elon Musk and Tesla elements of answer.