We scour the net for great ideas, so you don't have to
Claim this username to collect earnings from this post, and the portfolio!
TL;DR: Trucking will be the first viable business for autonomous vehicles. TuSimple ($TSP) currently has the best autonomous truck tech and is the closest to profitable commercialization.
Robotaxis are sexier and get all the media attention. But commercialization of robotaxis currently poses several technical, operational, and business challenges.
Robotaxis are most profitable in downtown environments. But those environments also pose the greatest challenges for an autonomous vehicle due to the large number of hard-to-predict agents. No company has been able to solve those yet.
Pick up and drop offs are also technically very challenging because, frankly, most Uber/Lyft pick up and drop offs in downtown environments are in illegal locations. Human drivers can get away with that. But a computer system cannot be programmed to perform illegal actions without the company being held responsible.
With Uber or Lyft, the drivers do much more than just drive. They purchase, store, clean, fuel/charge, and maintain their vehicles. With robotaxis, all of this operational load falls to the robotaxi company. Where are they going to store thousands of vehicles in downtown environments? How are they going to charge them and perform all of the other operational tasks? Even if this problem could be solved, it means that robotaxis will scale much more slowly than Uber or Lyft did.
The cost of robotaxis is currently extremely high due to the compute and sensor requirements. Even if they perform perfectly, it will take a long time before they are more profitable than vehicles with human drivers.
Further, recent regulation in the US has shown that the public support is behind the drivers, not the silicon valley companies. Driving for Uber and Lyft has been a lifeline to people who lost their jobs during the pandemic. So robotaxis are going to be viewed with hostility and likely backlash from the public and thus from governments.
Highway driving, though higher speed, is much more structured than downtown environments. In most ways it is much easier, as evidenced by current level 2 driver assistance systems that easily handle most highway driving. It does, however, require much longer range perception. We will return to that below.
Trucks already have all of the operational infrastructure. They are stored at depots. They have dedicated staff for their cleaning, maintenance, fueling, etc. Making trucks autonomous can simply leverage all of this infrastructure.
There is currently a shortage of truck drivers. Freight shippers want to ship faster (and not need to wait for drivers to sleep). Truck driving is not a fun or healthy job. It keeps people away from their families. And trucking is a $4T industry globally ($1T in the US).
Given the long lifespan of trucks and their ability to run for millions of miles, the cost of the autonomous compute and sensor system can be amortized much better than for robotaxis.
In short, the stars are all aligned for autonomous trucks today.
The players who started with robotaxis have built their entire tech stack around detailed but short range perception. Trying to convert that tech stack to trucking poses more challenges than one might expect. But since TuSimple has focused on trucking from the beginning (2015), they have built their stack to have 1000 meter perception - more than 4x other companies.
TuSimple also has been building up partnerships with the rest of the international freight ecosystem since the beginning. They have key partnerships with OEMs, Tier 1s, shipping companies, and part manufacturers in the US, Europe, and Asia. None of the other autonomous companies have that.
TuSimple is also building their Autonomous Freight Network across the US in multiple phases. Initially, only corridors with favorable weather will be autonomous. But as the technology improves to be able to handle heavy rain and snow, additional corridors can be easily flipped from manual to autonomous mode.
They definitely have ties to China. Their first investors were Chinese. So what?
This actually gives them an advantage vs pure US players: they have easy access to the entire trucking market in China.
Given all of the above, TuSimple is undervalued with its current valuation of about $15B. Some other self-driving companies are around $30B or more, despite facing all of the challenges mentioned.