Uber Shops For Self-Driving Partners Or Even Selling Its Project

Reports in the last two weeks have suggested the Uber
UBER
is looking to bring in external partners for their “advanced technology group” or even selling it outright to Aurora, a leading startup. Uber ATG is one of the most unusual — and certainly most controversial — self-driving teams. Strangely, Uber is the company best poised to monetize robocar technology, but also in some ways the one for which ownership of the technology is least essential.

Former Uber CEO Travis Kalanick felt it was essential. Uber is in the business of selling rides (AB5 debate notwithstanding) and many people feel that the business of selling rides will be taken over by the robotaxi. As such it is indeed essential that Uber be on top of that transition. That’s why he started the ATG project, feeling that Uber must own that tech. After lawsuits, the Otto acquisition and the fatality and huge costs with poor results, the new CEO has been less excited.

If, as many predict, the auto industry shifts to being more about selling rides than cars, then being Uber, with the world’s top brand in selling rides, is a good thing. Today Uber doesn’t own or operate their vehicles, so this would be a big shift, but it’s not an absolutely necessary shift. One can certainly imagine an Uber that, like today, simply connects you up with a robotaxi provided by some other owner or operator. In spite of there being no technological or major capital barriers to entry of creating an Uber competitor, only a few have pulled it off — the established brand and network size are that valuable.

While it would be an ideal position for Uber to have both the top brand in selling rides and control/ownership of top-tier self-drive technology, it would be sufficient if it has a strong partnership with the owner of the self-drive technology. So a deal with Aurora that included some special access to the technology by Uber could be almost as useful as owning the tech itself. Uber will choose between owning robotaxi fleets or licencing its technology to vehicle makers who want to put their vehicles into robotaxi fleets through Uber. To some extent, it may be able to do this even if it sells the tech to the right partner, or brings in other partners to the project.

On the other hand, Uber ATG’s tech is not well regarded. The fatality showed a team in disarray with poor performance and terrible safety culture. They’ve worked hard to fix that safety culture — they had little choice — but most imagine they are behind companies like Waymo, Cruise, Ama/Zoox and even Aurora. It’s unclear what they have to offer Aurora other than a team and may be a few technological pearls. Aurora has top talent, but has seen some setbacks as initial major partners like Volkswagen pulled away, and key investor Amazon
AMZN
purchased Zoox. A special relationship with Uber could provide a different and important market for Aurora’s technology.

The Robotaxi Business

Uber is obviously devoted to the taxi, and thus robotaxi business. That’s also the plan for Waymo and Zoox, and Ford/Argo and GM/Cruise have also said it’s their business, even though today they are car OEMs. Other car OEMs still contemplate a business of selling consumer cars with self-drive ability. Tesla, famously, hopes to do this but promises that once they have it, they will also operate a robotaxi fleet using the cars of private Tesla owners (plus off-lease Teslas
TSLA
that they retain the right to buy back.)

There is debate about what the best business model is, and whether the robotaxi market works financially in the near term. Over the long term, it should be possible to sell robotaxi service at scale for a price similar to, or even less than, the current cost of owning and operating your own car, much less than the $2/mile typical cost of rides like Uber. In the short term, however, this is difficult due to the very high costs of software and hardware development and the more limited service areas.

A company like Uber has another unique advantage here. They have a large number of customers ordering rides. If the customer asks for a ride that can be serviced by the robot, they can send the robot. If the customer wants to go further afield, they can just send a regular Uber driver. They have the option of pricing those differently, or pricing them the same and subsidizing the human driver rides with extra profit on robot rides. Customers just want to go where they want to go, and will like a company that can always make that happen, which a pure robotaxi company can’t.

For Uber, if their ATG self-driving project fails, they are still Uber, and they still have a shot at managing rides in other people’s cars. It will be harder to maintain that position because it will be dealing with large companies rather than private individuals, but it is still possible. If Waymo, Cruise, Argo or Aurora fail to make their system work, they are simply doomed as a company. Uber can take advantage of this in making partnerships.