Tesla may ditch radar altogether in its bid to achieve fully autonomous driving, CEO Elon Musk has claimed in the company’s latest earnings briefing for analysts and investors.
The global leader in electric vehicles has come under intense scrutiny of late, with the death of two men in Texas prompting many in the media to call into question its Autopilot technology. Tesla angrily rejects those claims, pointing out the reports that Tesla Model S involved was operating on Autopilot are clearly false, and claims it took four hours to put out the ensuing fire have also been debunked.
Musk has previously said that LiDAR (light-emitting radar), which underpins the self-driving hopes of many other stakeholders looking to offer autonomous driving, is too energy-intensive.
Tesla is now adding radar to its list of possibly superfluous technologies, saying in its Q1 2021 earnings report today (Australia time) that, “We believe that a vision-only system is ultimately all that is needed for full autonomy.
“Our AI-based software architecture has been increasingly reliant on cameras, to the point where radar is becoming unnecessary earlier than expected,” it says. “As a result, our FSD team is fully focused on evolving to a vision-based autonomous system and we are nearly ready to switch the US market to Tesla Vision.”
Expanding on the subject of AI, autonomous driving and Tesla’s “Dojo” neural network that is being trained to process the vast amounts of visual imagery being transmitted to Tesla via a million electric cars on the road, and twice that expected by the end of 2022, Musk said that relying on multiple cameras has “an extremely high level of safety.”
“Think of Tesla, as much as an AI, robotics company, as we are a company or an energy company,” he said.
“We couldn’t find a powerful enough neural net for computers that we designed and built our own. The software out there was really quite primitive for the task so we have been developing what we think is the most advanced real-world AI in the world.
“And then it sort of makes sense to this is what needs to happen because the road system is designed for a neural net – our brain …. and it’s an entire system designed for vision with a neural net computer – eyes and a brain.
“And so, if you have a system which has very good eyes in all directions at once. with three focal points forward, it never gets tired, its never texting, it has redundancy, and its reaction time is superhuman, then it seems pretty obvious that such a system would have an extremely high level of safety.”
While Tesla may be ditching radar in its exterior sensors, it may be adding them inside the car.
According to Teslarati, Tesla has permissions to install high-tech millimetre-wave sensors inside its vehicles. Tesla and five other companies have received a “grant of waivers” allowing the company to install the sensors inside the car, with Tesla granted permission to use the 60-64 GHz frequency.
The sensors could be used for purposes such as detecting if there is a child inside the car or detecting an intrusion into the cabin.
It could also be used to monitor drivers who are using Autopilot or testing the company’s full self-driving (FSD) beta software.
Tesla reportedly recently removed some drivers from its beta program as they did not remain attentive while driving with FSD active.
Bridie Schmidt is lead reporter for The Driven, sister site of Renew Economy. She has been writing about electric vehicles since 2018, and has a keen interest in the role that zero-emissions transport has to play in sustainability. She has participated in podcasts such as Download This Show with Marc Fennell and Shirtloads of Science with Karl Kruszelnicki and is co-organiser of the Northern Rivers Electric Vehicle Forum. Bridie also owns a Tesla Model 3 and has it available for hire on evee.com.au.