The company’s rollout of its new driverless cars has gotten off to a wobbly start – and rival Waymo remains well ahead
After years of promising investors that millions of Tesla robotaxis would soon fill the streets, Elon Musk debuted his driverless car service in a limited public rollout in Austin, Texas. It did not go smoothly.
The 22 June launch initially appeared successful enough, with a flood of videos from pro-Tesla social media influencers praising the service and sharing footage of their rides. Musk celebrated it as a triumph, and the following day, Tesla’s stock rose nearly 10%.
What quickly became apparent, however, was that the same influencer videos Musk promoted also depicted the self-driving cars appearing to break traffic laws or struggle to properly function. By Tuesday, the National Highway Traffic Safety Administration (NHTSA) had opened an investigation into the service and requested information from Tesla on the incidents.
Let me tell you how thrilled we all are to have a new hazard added to Austin streets.
"LIDAR is so lame! Who needs to be able to accurately map their surroundings, when our cars can get confused by a shadow passing across one of their lenses, just like humans do? It's so much more efficient to be blinded by fog, or low light conditions."
Our eyes are not perfect organs so why pretend like they are? Our eyes fail us:
when it’s too dark
when it’s too bright
when there’s fog
when there’s too much rain and snow
when there’s glare from the sun
when there’s obstructions
when there’s sensory overload
when there’s something covering our eyes like dirt and mud
when we can only see on the visible spectrum
Why wouldn’t we want more incoming data to account for these shortcomings? Optical-only vision-based solutions are incomplete because our eyes are incomplete. I can’t see that a car is stopped dead in the road 10 feet ahead of me in thick fog, but an advanced set of telemetry sensors can. My eyes are not better than the scores of technology we’ve built over the past few decades and I’ve been practicing with them for 46 years. Give me a helmet that includes LIDAR and infrared and night vision and sonar and telemetry from a satellite and GPS and weather tracking and god knows what else and I’ll be much less likely to rear end that car in the fog. We humans invent technology to make up for our shortcomings, so why go with the idea of “if it’s good enough for biological evolution it’s good enough for these multi-ton contraptions we have hurtling down highways next to each other several metres apart at 100 km per hour every second of every day?” It sounds ludicrous on its face. We can choke on a peanut because our swallow tube is the next to the breathing tube ffs. We can do better.
Yeah, the "appeal to nature" fallacy that Musk is using is really dangerous. I want the best tools that are practical for the job, since this has actual life and death consequences. A mix of sensors that work together and covers each others limitations.
With good integrated design, the LIDAR could be practically invisible. So weird to think the average person actually would care about the details of how something works, or maybe Elon Musk just literally cannot imagine a car that uses LIDAR without it having a big assembly on the roof.
This has nothing to do with aesthetics and everything to do with reducing production costs. The sensors themselves add to costs, but then the data processing for all of this data adds to the computational needs of the system, sensor installs requiring calibration and flexibility for changes in sensors add costs. The list goes on.
Musk and Tesla are standing firm on their position that LiDAR is needless and that others will follow their lead here, but so far in practice the lack of LiDAR in Tesla vehicles has shown to be a serious disadvantage compared to other systems.
Yep, I used to work for a robotics company that makes self-navigating floor cleaning robots. The LIDAR is completely integrated into the housing of the robot in both the old designs and the new one; it's barely noticeable.
I have a theory on this. They use a lot of ML for FSD and Ap. That a radical redesign would change the model enough that they would have to retrain from scratch. So they keep the facelifts extremely minor .
Always interesting to see the view of the degree of Elon Musk's involvement in his companies' decisions swing depending on whether the outcome is good or bad.
What I mean is that when Musk-owned companies have successes people are very often quick to accuse him of "just hiring smart people" or "just buying a successful company." It's only when those companies have failures that he gets credit for being hands-on in their design decisions.
Don't get me wrong, I think Elon Musk is a pretty terrible person both in terms of his personality and his politics. But pretty terrible people can nevertheless be smart and make good engineering decisions. Just look at von Braun as a prime example.
The company has not released information on whether, or how long, it has spent mapping out or testing the driverless technology on Austin’s streets.
That's being too nice, the CEO is clearly proud of not mapping cities, as seen in the tweets. The journalist should call it out explicitly. "They are probably not mapping as suggested by the CEO's public posts." That's not too much of a leap.