3 Autonomous Driving Technologies Investors Must Know About in 2022

While Tesla is taking a camera-based vision approach to developing its self-driving vehicle technology, Waymo and other leaders in the space are using high-definition maps to build a fully autonomous service. In this Backstage Pass clip from "The AI/ML Show" recorded on Jan. 12, Motley Fool contributor Trevor Jennewine outlines and explains the pros and cons of these different technologies. 

Trevor Jennewine: I wanted to read a quote from Andrej Karpathy going back to why Tesla got rid of RADAR, why they're not using LiDAR. This is from Andrej Karpathy back in June 2020 and he said Waymo and many others in the industry use high-definition maps. That means you first have to drive some car that pre-maps the environment. You have to have LiDAR with centimeter-level accuracy and you are on rails. You know exactly how you're going to turn in the intersection, you know exactly which traffic lights are relevant to you, you know where they are positioned, you know everything.

This slide right here the graphic is provided by ARK Invest and it shows the same concept. Tesla, their camera-based, their vision approach, they're alone in taking that approach, and the reason they do that is because they believe it's more scalable. You can see the colors of the bar represent the difficulty level, so the pink, red, orange. Tesla's approach is very difficult, but if you look at along the righthand side, it's also the most scalable approach and that makes sense.

What Waymo and other companies are doing is they're driving these cars around an environment. Waymo already has a fully autonomous service that they offer in Phoenix, Arizona, they're rolling it out to testers in San Francisco, and the way they do that is they drive their cars around they're using LiDAR with centimeter-level accuracy to build these HD maps, and so the car really is on rails. You couldn't take Waymo's cars into a city that it hasn't mapped and expect them to perform the same way they wouldn't be able to do that.

There are pros and cons to this Waymo is already up and running. That's great. They're getting their name out there. That's a definitely pro for Waymo. However, they're going to have to continue to map all of the roads that they operate on, and if anything changes, any construction or anything about the road itself changes that's going to have to be remapped and the maps will have to be maintained over time.

Tesla's approach is much more scalable. It's going to be more difficult to get there, but if they do get there and they have neural networks that are able to power the car as soon as they have that, they can send it over to all their cars and all their cars will be able to drive autonomously immediately. I think that I wanted to really emphasize that point. There are a lot of different approaches here and Tesla falls into its own camp. Most other companies are using that HD mapping with LiDAR, RADAR to support the cameras.

ARK Invest is obviously very bullish on Tesla. I put this next slide in here. This is from an IT consultancy, Guidehouse, they feel the exact opposite. They like Waymo's and Nvidia, Cruise. They are ranking them as leaders and Tesla is very at the far the bottom left. Indicating a worse execution and a worse strategy. One of the things I think is interesting about this space is that there are smart people on both sides of the argument and I think that makes autonomous driving very interesting. Personally, my take is that if Tesla can solve the problem without using RADAR and LiDAR once it has those neural networks, the fact that we'll be able to instantly scale, I think that's valuable.