BY CHRISTOPH MEYER
ME 302C - The Future of the Automobile - Mobility Entrepreneurship is a Stanford course taught by Reilly Brennan and Stephen Zoepf in Spring 2017. This course will feature a series of guest speakers across the spectrum of the mobility industry, with Pearl, Turo, and Lyft featured among others.
In order to provide an inside view to mobility enthusiasts not enrolled in the course, I will be publishing a weekly post on the CARS blog. These posts will feature a high level summary of a brief discussion with each of the guest speakers. While the content shared in the classroom will be kept confidential and off-the-record, I will interview them for 15-20 minutes afterwards. I may probe them on topics that were covered in class but the interviewee will have the discretion as to how they answer. The goal will be to learn more about their companies as well as delve into the speakers’ views on general trends in the industry and predictions of where things are headed.
Keep your eyes out for these weekly updates - full speed ahead!
Interview with Stefan Seltz-Axmacher-CEO and founder of Starsky Robotics
I really enjoyed Stefan’s candor during class and in our interview. He provided his perspective on a broad range of topics and showed why he is in charge of one of the companies seeking to lead the way forward in trucking. Throughout the conversation, it was evident that Stefan is guiding Starsky to think and compete differently, embracing alternative strategies and rejecting industry trends. While the space may be nascent, conventions and principles have already begun to form. Some approaches have shown more promise early on than others, leading to some converging. This has also been exacerbated by many of the leading engineers and visionaries jumping around between companies, startups, and competitors. After my interview with Stefan, I am inclined to believe this merging of thoughts is happening too early. In solving such complex problems, outside-the-box thinking and different approaches (such as using remote control) will be needed to ensure the industry and society (which Stefan commented on repeatedly) arrive at the best possible outcome.
Q: By incorporating remote control, you’ve taken a differentiated approach - how do you think about being different?
“I don’t think there is a lot of value in the orthodoxy of other players. This space is so new. Maybe the big players or those that did the DARPA Grand Challenge didn’t do it the way we did. But they also haven’t brought anything to market. People have questioned why we don’t use LIDAR and think we are crazy. But those who are using it haven’t brought anything to market yet either. For a space where nobody has really done anything proven, there’s a lot of dogma.”
Q: How do you think about some of the big obstacles facing your approach, such as latency and cyber security?
“There are a lot of other ways around these issues. Usually in startups, there’s a general truism: the dynamic of a business founder telling all the engineers what to do and build. This is different: it is an industry where engineers tell all the engineers to make things work. People are trying to do everything at the same time. There are a lot of places in the industry where people have tried to do everything at once and that just doesn’t work. Cruise is doing really well by focusing only on San Francisco. They just focused on one place, while Google is trying to work everywhere. My view is that you have to perfect one area and build out from there. Even perfecting a narrow area is hard, but making it work all the time is impossible.”
Q: Do you think that remote control is an interim or long-term solution?
“I think it is a pretty long term solution. In driving, 90% is going straight then 9% is turning. Each additional ‘9’ of reliability is an order of magnitude more of complexity. Even if we get out to the seventh ‘9’ (99.99999% reliability), some people won’t be comfortable. A lot of other big players recognize this and have started filing patents for tele-operators. The idea that a computer controls a moving vehicle all the time is one of the stupid parts of the industry’s dogma. LIDAR is a huge misperception of the industry. It’s fairy dust. It sounds cool but it’s not built automotive grade. And, software cycles faster than hardware. So I really think LIDAR is unlikely to be the final, long-term solution.”
Q: With players approaching the problem from different perspectives and different sub-markets, how do you think the industry evolves?
“The biggest fish to catch here is self-driving taxis. That’s huge. Self driving taxis are a $10 trillion per year industry. If Uber focuses its engineering on trucks rather than taxis, they will lose. Most people are much more focused on the taxi side of the equation. A lot of the taxi companies that don’t know what they’re doing will go into trucks once they’ve lost the battle over taxis. Trucks can quickly produce more miles, but in a different environment. It’s very possible that truck companies like us will get more mileage data than the big players (Waymo, etc) very quickly. Starsky is not focused on taxis and we don’t want to be. The opportunity in trucks is still massive and we believe we can do very well here.”
Q: How do you see the dynamic of robots taking away jobs playing out longer term?
“Keynes predicted years ago that by our time, we would only work 4 hours per week. That clearly hasn’t happened. We have found other things to do with our time. I think there is a strong case for basic income. Our investors think about this a lot and I think this will be a likely solution. The current nature of work is a relatively new idea - it hasn’t been like this forever and it probably won’t stay this way in the future. Robots and AI will make the marginal costs of all goods close to zero. I also think nobody has ever done a re-training the right way.
This is one of the reasons that the winners in this space need to be good people. It really matters who wins this race - they can have a huge impact on society. A lot of companies think of themselves very highly and look down on the people and the parts of the economy they are looking to replace. We do not. The outcome of our industry will have a huge impact on society.”
Q: When do you think autonomous vehicles will go mainstream?
“It’s a lot faster than people think - especially for individual companies that are focusing on narrow use cases. Our company is talking about taking people out of trucks by end of the year. People have predictions for 2020 and beyond but a lot of people are closer than that. I predict that we will first see it become mainstream in California. Innovation tends to first appear in society closest to where the innovators are. In this industry, that’s California and particularly Silicon Valley.”
Q: What is the biggest obstacle to the evolution of this industry?
“The biggest risk is one company doing something bad that will negatively impact and slow down the rest of the industry. A serious accident could have major implications for the company involved, potentially ending their existence. There was a similar issue in the 1920’s-1930’s, when companies started making vaccines. Any vaccine that hurt people led them to be sued out of business. Of course vaccines were a good thing but nobody was incentivized to produce them. It was just too risky. Government needed to step in, help, and change the rules of the game. When this industry plays out, I’m either going to be on the level of Rockefeller or in jail. There is just such huge risk. Most older people and companies won’t take this risk and thus aren’t seriously in this industry. You have to have a huge, almost crazy risk tolerance and that’s why the people changing this industry are young people and early stage companies.”
Q: How do you think the evolution of the industry affects how people will live in the future?
“I grew up in the suburbs of Washington DC and thought it was boring, so I’m not sure that autonomous vehicles will make the suburbs more exciting. But, in America, people want choices and the option to buy cheaper homes and have a big yard. Autonomous vehicles will make that option easier by opening up land farther out and decreasing the time cost of driving. On the other hand, land in cities will also be freed up. Parking garages and other places will become available for housing and ease some of the space pressure, potentially lowering price. I am not sure which way it will go - it could go either. What is certain is that it will give people more choices.”