During Tesla’s earnings call in January, an analyst asked Elon Musk about his belief that Tesla would achieve Level 5 range – lingo for a car capable of driving in all situations – by the end of this year. the year.
“I’m confident based on my understanding of the technical roadmap and the progress we’re making between each beta iteration,” Musk said.
But six weeks later, Tesla’s Autopilot software director CJ Moore contradicted Musk in a March meeting with California regulators. This is according to a note obtained by transparency from the Plainsite site via a request for access to information.
According to LinkedIn, Moore has worked at Tesla for almost seven years and became director of Autopilot software in 2019. He was one of four Tesla representatives to participate in a conference call with officials from the California Department of Motor Vehicles on March 9.
“DMV asked CJ to process Elon’s L5 capacity message from an engineering perspective by the end of the year,” reads a note from a DMV official. summing up the meeting. “Elon’s tweet doesn’t match technical reality by CJ. »
Someone – presumably the California DMV – seems to have tried and failed to craft that last sentence. This part of the letter appears as white space in the PDF, but it can be copied and pasted into another document.
It is not known which tweet the memo refers to. I looked at Musk’s tweets in the weeks leading up to March 9 and couldn’t find any on the timeline to achieve level 5 autonomy.
But the January earnings call wasn’t the only time Musk predicted Tesla would achieve Level 5 autonomy by the end of 2021. In a December 2020 interview, Musk said he was “extremely confident” that Tesla vehicles would reach Level 5 within a year. .
The DMV memo then goes on to summarize comments from Tesla representatives:
Tesla is currently at Level 2. The driver interaction ratio would need to be on the order of 1 or 2 million miles per driver interaction to move to higher levels of automation. Tesla indicated that Elon extrapolated the improvement rates by talking about L5 abilities. Tesla couldn’t say if the improvement rate would hit L5 by the end of the calendar year.
It’s unclear how Tesla defines ‘driver interaction’, but it appears to be an extremely high bar. The last time I watched the FSD beta videos closely in real life situations, I saw the software make a number of errors within a few hours of video. Some recent videos continued to be shown frequent mistakesSo it seems unlikely that Tesla will reduce the error rate to once per million kilometers in the next eight months.
This isn’t the first time Tesla has seemed to say one thing to California regulators while Musk has told the public another. Letters to the California DMV in November and December asserted that the beta version of the FSD “will continue to be an SAE Level 2 Advanced Driver Assistance feature”, not a fully stand-alone system. Tesla said work on “true stand-alone features” would begin at a later date.
FSD has been absent for one to two years for five years
This should all sound familiar to longtime Tesla watchers. Two years ago, Elon Musk confidently predicted that by the end of 2020, Tesla’s software would be sophisticated enough to allow Tesla vehicles to function as fully autonomous robotaxis. Needless to say, nothing like this happened.
Three years ago this, in June 2016, Elon Musk was asked how long it would take to make fully autonomous vehicles.
“I would really view autonomous driving as a solved problem,” Musk said. “I think we’re basically less than two years away from full autonomy. “
Four months later, Tesla unveiled a new version of its Autopilot hardware. The company billed the new vehicles as being capable of fully autonomous driving and began accepting payments for Tesla’s future full autonomous driving software. Musk predicted that a Tesla would be able to drive coast to coast “without the need of a single touch” by the end of 2017.
CJ Moore was also not the first autopilot executive to be uncomfortable with his boss’s grandiose promises of full autonomous driving technology. Sterling Anderson was the head of Tesla’s autopilot program when Tesla started selling the full self-driving package in 2016.
“The announcement upset some engineers because they thought the product released was not designed to be stand-alone,” according to a 2017 article published in the the Le journal Wall Street. When an employee asked Anderson about the move, he reportedly replied that “it was Elon’s decision.”
Anderson resigned shortly thereafter and ended up co-founding standalone startup Aurora. Musk’s decision to market Autopilot as a “fully autonomous” system “was a factor in the decision of Mr. Anderson and at least two other engineers to leave the company,” according to the Journal.