Stories related to autonomous driving tech have been building up, so I’m going to run through a few of them right here.
Comma.ai Could Buy Automaker Someday
The first one is surprising but logical, if you’ve been following the autonomous vehicle industry and have faith in Comma.ai’s approach.
A little more than 5 years ago, George Hotz, who was the first person to hack an iPhone when he was a teenager and worked at several large tech companies after that, decided to build an aftermarket autonomous driving system out of his garage. (After all, what better place to launch a tech giant?) I was a moderator at a conference Hotz was speaking at in 2017 in Paris, France, and joined his session in the audience. (He missed his flight — or “missed his flight” — from California and had to livestream it in, almost as if he was 3 years early to a pandemic lifestyle.) Hotz claimed that Tesla, with its Autopilot/Full Self-Driving approach, was essentially the Apple of this burgeoning industry, but that Comma.ai would be the Android of it. More recently, in mid-2019, Hotz was on Lex Fridman’s Artificial Intelligence podcast/YouTube show, and that may well be the most interesting discussion of autonomous driving tech I’ve watched, so I highly recommend it. Hotz continues to claim that Tesla will be first with true self-driving cars (outside of geofencing) and Comma.ai will be second. With that in my, he indicated yesterday on Twitter that Comma.ai may well buy an automaker one day.
comma is more likely to acquire a car company (after we solve self-driving cars) than a car company is to acquire us.
— comma (@comma_ai) January 9, 2021
Again, if you believe that Tesla and Comma.ai are taking the right approach to this difficult challenge, it does seem completely logical that the latter would one day buy an automaker. Hotz has been steadfastly unwilling to forecast when the problem will be solved enough to truly put self-driving cars/robotaxis on the road, but he’s convinced it will happen and Comma.ai will trail Tesla by a couple of years. I highly recommend his discussion with Lex as well as his 2017 comments at the Autonomy conference in Paris:
Waymo: (Tesla’s) “Self-Driving” Terminology Leaves False Impressions
Waymo (which spun out of Google and became a subsidiary of the tech giant) is widely referred to as the leader in this space. Of course, the never-ending debate is whether Waymo’s approach or Tesla’s approach is better. We shall see. In the meantime, Waymo does actually have robotaxi service (Waymo One) in operation in Phoenix, Arizona, and I expect it will launch such service somewhere in California soon.
Before getting to Waymo’s news on terminology, here are a couple of interesting tweets from Waymo CEO John Krafcik:
To get people where they need to go, safely & easily, we'll offer Waymo One in cities & communities around the world. You can use it in PHX today.
To move packages & other things, Waymo Via will connect shippers & customers.
That's the plan. https://t.co/6F2bu6tvEs /2
— John Krafcik (@johnkrafcik) December 28, 2020
That is sure to kick off more debate — the matter of real-world driving miles versus simulated driving miles is great for that.
Soon after those notes of the company’s autonomous driving leadership, Waymo put out a statement indicating that it doesn’t think “self-driving” is an appropriate term in this industry and has stopped using it in its communications. It wants to use “more deliberate language, referring to our fully autonomous driving technology, and no longer referencing ‘self-driving.'” Personally, I don’t see a notable difference between the terms — they are synonymous. “Semi-autonomous” is different, but who is “autonomous” and “self-driving” different in substance? While I wait for someone to enlighten me, let’s move on.
It appears that the bigger point here is that Waymo wants to differentiate itself from Tesla, which of course offers a “Full Self-Driving” suite of driver-assist features that should eventually (but does not yet) provide fully autonomous robotaxi-level service. Waymo writes:
“We’re hopeful that consistency will help differentiate the fully autonomous technology Waymo is developing from driver-assist technologies (sometimes erroneously referred to as ‘self-driving’ technologies) that require oversight from licensed human drivers for safe operation.”
If that doesn’t make its criticism of Tesla and desire to be seen as doing a different thing clear enough, though, here’s more from the company blog post:
“Today, the Waymo Driver makes billions of decisions each day as it safely moves people and goods to their destination in fully autonomous mode.
“This is more than just a branding or linguistic exercise. Unfortunately, we see that some automakers use the term “self-driving” in an inaccurate way, giving consumers and the general public a false impression of the capabilities of driver assist (not fully autonomous) technology. That false impression can lead someone to unknowingly take risks (like taking their hands off the steering wheel) that could jeopardize not only their own safety but the safety of people around them. Coalescing around standard terminology will not just prevent misunderstanding and confusion, it will also save lives. To that end, we’ve renamed our public education campaign as Let’s Talk Autonomous Driving.”
This debate around Tesla’s terminology has been going on for years. As someone with a Tesla Model 3 with the Full Self-Driving suite, I feel like the explanations and how the system responds to you make it clear that this is only driver-assist technology that aspires to become fully self-driving tech at some point in the future. The car won’t even let you take your hands off the wheel for long. However, I’ve also had some ask me if the car can completely drive itself. My opinion is that people who don’t have Teslas are concerned about the technology but that I’m yet to meet someone with a Tesla — especially someone who dropped several thousand dollars, or $10,000 now, on the FSD suite — who thought the tech could do more than it can. Chime in with your own 2,000 cents down in the comments.
Xpeng Adds Lidar
Xpeng, a young smart electric vehicle startup out of China, is adding lidar technology to its XPILOT semi-autonomous driving hardware in 2021 production models, making it the first company in the world to be selling mass-produced EVs that include lidar.
Xpeng is getting its automotive-grade lidar from Livox. “Livox has customized its Horiz sensor (the automotive-grade version of its Livox Horizon) for Xpeng, not only meeting automotive-grade requirements but also delivering cost efficiency and reliability for production models. The implementation of lidar in XPILOT architecture – Xpeng’s autonomous driving system – will further enhance XPILOT’s safety as well as the ability to cover a comprehensive range of driving scenarios.
“Since its founding in 2016, Livox has focused its R&D efforts on enabling cost-efficient mass production of lidar solutions, aiming to overcome the three bottlenecks in the lidar industry: price, scalability and reliability. Livox’s automotive-grade solution boasts a range of industry-leading standards in terms of detection range, FOV (Field of View) and point cloud density.”
More details can be found here. Will other automakers add lidar to normal consumer electric vehicles in 2021? We’ll see.
Disclosure: I own shares in Tesla (NASDAQ:TSLA) and Xpeng (NYSE:XPEV). Nothing above is investment advice, and I do not provide investment advice, not even to my sister.