Tesla Lays Off About 200 Autopilot Workers, Most of Them Hourly

(Bloomberg) — Tesla Inc. laid off hundreds of workers on its Autopilot team as the electric-vehicle maker shuttered a California facility, according to people familiar with the matter.

Most Read from Bloomberg

Surprisingly, the majority of those who were let go were hourly workers, said the people, who asked not to be identified discussing private information. As recently as last week, Chief Executive Officer Elon Musk had outlined plans to cut 10% of salaried staff but said he’d be increasing hourly jobs.

Teams at the San Mateo office were tasked with evaluating customer vehicle data related to the Autopilot driver-assistance features and performing so-called data labeling. Many of the staff were data annotation specialists, all of which are hourly positions, one of the people said.

About 200 workers were let go in total, according to the people. Prior to the cuts, the office had about 350 employees, some of whom were already transferred to a nearby facility in recent weeks.

Tesla didn’t immediately respond to a request for comment.

Tesla is trimming its ranks after a surge in hiring in recent years. The company, now based in Austin, Texas, had grown to about 100,000 employees globally as it built new factories in Austin and Berlin.

See also: Elon Musk Sounds Off on Recession Risk, Twitter Deal and Trump

Musk caught workers by surprise earlier this month when he said layoffs would be necessary in an increasingly shaky economic environment. He clarified in an interview with Bloomberg that about 10% of salaried employees would lose their jobs over the next three months, though the overall headcount could be higher in a year.

The EV market leader’s downsizing efforts have focused on areas that grew too quickly. Some human resources workers and software engineers are among those who have been laid off, and in some cases, the cuts have hit employees who had worked at the company for just a few weeks.

Those affected by the latest move worked on one of the higher-profile features in Tesla vehicles. In job postings, Tesla has said that labeled data is the “critical ingredient for training powerful Deep Neural Networks, which help drive the Tesla vehicles autonomously.” Staffers in Buffalo, New York, and San Mateo spent hours labeling images for cars and the environment they navigate, such as street signs and traffic lanes.

In Buffalo, Tesla has continued to expand its Autopilot data-labeling teams, a person familiar with the matter said. But staff at that location, who are doing the same role, are paid a lower hourly rate than in San Mateo, the person said.

Tesla’s shares fell less than 1% in late trading. The stock tumbled 34% this year through Tuesday’s close, compared with a 20% decline in the SP 500 Index.