The AV Industry Searches for a Near-Term Business Case

October 21, 2019

Bradley Berman

In July 2019, an impressive lineup of the world’s leading autonomous vehicle (AV) companies gathered in San Jose for the TechCrunch Sessions: Mobility 2019 conference. Technologists, executives and transportation experts took to the stage of the California Theater, a beautifully restored 1920s-era motion-picture house. The venue has a long legacy of presenting fantastical movies, operas, and vaudeville shows. However, the day-long, back-to-back 20-minute conversations on stage struck a note of hyper-realism.

Executives from Aptiv, Aurora, Mobileye, Uber, Waymo and other companies mostly dodged questions about when self-driving technology will reach maturity. Instead, the dominant theme was the search for practical, near-term and profitable ways to solve real mobility problems. The presenters on the TechCrunch stage who seemed the most comfortable with the shift in message from revolution to evolution are working on well-defined use cases and business models.

Make money by doing less

Daniel Laury, co-founder and chief executive of Udelv, said his company is “perhaps the most successful autonomous driving company in the world right now.” How can Laury make this claim for an AV delivery company that is practically unknown? “We’re not just focusing on autonomous driving. We’re also focusing on the business,” he said. “To answer the needs of retailers and shippers, we created the robotic delivery system that works for retailers. It’s a complete end-to-end solution, which I think is the future for the next five to 10 or 15 years of autonomous driving.”

Another clear use case is fixed-route shuttle service. Alisyn Malek, co-founder and chief operating officer of May Mobility, said companies that promised in 2016 and 2017 to launch robotaxi services by 2018 are still not in service. She explained that mapping an entire city is a lot harder than serving a neighborhood with high utilization. “We’re focused on building a business,” she said. “What’s the least amount we can do to make the most amount of money? There’s a lot of money in shuttles.”

May Mobility will start with low-speed electric shuttles with a maximum speed of 25 miles per hour. This approach makes it easier to overcome regulatory hurdles while building business relationships and operational experience. May Mobility will potentially develop a different type of vehicle with expanded capabilities, but only after learning what type of self-driving vehicle is best for users.

Iterative vehicle design is also the name of the game for Nuro, another autonomous-vehicle delivery service. Nuro’s cute first-generation robot – about half the width of a Toyota Corolla – was custom-built to handle local errands, such as delivering groceries or dry cleaning. Dave Ferguson, co-founder and president of Nuro, said that 43% of the 400 billion personal vehicle trips a year are for running errands. He wants Nuro to reduce those trips.

Voyage, a company providing shuttle service to retirement communities, is taking a similar approach. “The beginning of Voyage was very much informed by finding a customer first and working backward to figure out what technology those folks needed,” said Oliver Cameron, co-founder and chief executive of Voyage. “I think all too often in self-driving cars, you start with technology,” he said. “That’s crazy.”

Michael Granoff, founder and managing partner of Maniv Mobility, an Israel-based venture fund, was a panelist for TechCrunch’s session about the role of venture capital in future mobility. He drove the point home about solving real needs instead of making revolutionary promises. “One of the truest cliches in businesses is to fall in love with the problem and not the solution. Think about what’s the problem for you in your own mobility life and for your peers. Figure out how to solve that.”

Build toward the mass market

A new focus on near-term mobility needs doesn’t mean that AV companies will take their eyes off the long-term promise. Several panelists at the TechCrunch event discussed how vehicle designs were shifting from a small fleet of custom prototypes to vehicle platforms that can be produced at scale. For example, the evolution of Nuro’s second- and third-generation delivery bots was focused on increasing the potential for mass manufacturing.

Waymo has deployed five different platforms over a decade, including the Toyota Prius, Lexus RX450h SUV, Chrysler Pacifica minivan, Jaguar I-Pace electric crossover and its iconic Google car. The company is now using the fifth generation of its self-driving hardware, in which manufacturing at scale is an increasing priority. “Unit economics is not something you fix after the fact,” said Dmitri Dolgov, Waymo’s chief technology officer.

Resetting expectations

Safety was also a recurring theme at the event, which took place a mere 16 months after one of Uber’s autonomous test vehicles killed Elaine Herzberg in the first recorded pedestrian fatality involving an AV. Waymo’s Dolgov explained that there had been significant technological gains in the past two years. “Cars that are completely driverless with no test driver behind the wheel are operating on public roads,” he said. But Dolgov cautioned, “It’s not [just] one or two things that remain to solve everything.” He pointed to the need for more powerful sensors with increased range and resolution, as well as cheaper compute software and sensor fusion to classify cars, pedestrians, cyclists and roadway objects.

Karl Iagnemma, president of Aptiv, a tier-one supplier specializing in automated technology, championed his company’s partnership with Lyft. In that service, he said, more than 50,000 rides in a self-driving car have been given in Las Vegas with “no at-fault accidents.” But he also struck a cautionary note, especially about overpromising a zero-accident future. “I’m actually optimistic that if we’re responsible as an industry about how we communicate about the technology and don’t create false perceptions on what it is and isn’t, then we’ll see good societal acceptance even when crashes occur,” Iagnemma said. And a big part of the true picture, according to Iagnemma, is a “gradual change” from human to automated driving.

---