Likely lost in the coronavirus news onslaught last week was the demise of Starsky Robotics. A startup failing is not news; most such businesses fail. Still, when co-founder Stefan Seltz-Axmacher took to Medium last Thursday to announce the closure, he dove into great detail on what happened, and where he sees the autonomous vehicle space.
Here’s the highlight: If you’re not on the autonomous vehicle bandwagon yet, you have plenty of time.
“From my vantage point, I think the most likely line of human equivalence is L3 [autonomy] which means that no one should be betting a business on safe AI (artificial intelligence) decision makers,” Seltz-Axmacher wrote. “The current companies who are will continue to drain momentum over the next two years, followed by a few years with nearly no investment in the space, and [hopefully] another unmanned highway test [in] five years.”
Seltz-Axmacher went through the issues Starsky had but also some of the broader issues facing autonomous vehicle development.
Starsky was somewhat atypical in what it was building. The company focused on building an autonomous system that allowed a truck to drive unoccupied on roadways, but it would always be monitored by a remote driver. That driver would use video cameras and vehicle controls such as a steering wheel to take control of the truck as needed. The driver would also drive the vehicle from the control room during the last mile of deliveries, similar to a video game. That means that while there would be no physical driver in the vehicle, a driver would remain in control.
But, unlike autonomous startups like TuSimple, Ike, Plus.ai and Waymo, Starsky was never able to attract the big funding necessary to develop the technology. Seltz-Axmacher acknowledged that the Starsky team didn’t understand what venture capitalists (VCs) were looking for in an autonomous investment.
“It took me way too long to realize that VCs would rather [invest in] a $1B business with a 90% margin than a $5B business with a 50% margin, even if capital requirements and growth were the same,” he wrote.
Crunchbase data shows that TuSimple has raised around $298 million in the past six rounds of funding and self-driving startup Plus.ai has raised around $200 million in its past three rounds of funding. Ike has raised $52 million in Series A funding, and Embark, Waymo and others have been able to attract significant funding as well. Starsky attracted just $20.3 million in funding in its three years.
“It took me way too long to realize that VCs would rather [invest in] a $1B business with a 90% margin than a $5B business with a 50% margin, even if capital requirements and growth were the same.”Stefan Seltz-Axmacher, Starsky Robotics co-founder
Seltz-Axmacher wrote that plenty of brokers are willing to place freight on autonomous vehicles, but what limits the opportunity is the safety aspect.
“In January of 2019, our head of safety, our head of PR, and I gathered in a conference room for a strategy session. The issue: how could we make safety seem exciting enough to cover,” Seltz-Axmacher wrote. “A month earlier we had publicly released our VSSA [voluntary safety self-assessment], a highly technical document that detailed how we decided to approach safety. We had pitched it to a particularly smart reporter, but instead of covering it in detail, they mostly wrote about teleoperation. We left the meeting in a fluster — we couldn’t figure out how to make safety engineering sexy enough to garner its own reporting.
“And we never really figured out how,” he added.
Starsky worked hard at documenting and fixing system failures, Seltz-Axmacher noted, but he said investors don’t see, or value, that work.
“The problem is that all of that work is invisible. Investors expect founders to lie to them — so how are they to believe that the unmanned run we did actually only had a one in a million chance of [a fatal] accident? If they don’t know how hard it is to do unmanned, how do they know someone else can’t do it next week?” he wrote.
What did get investors’ attention were competitors who invested in AI features.
“Decision makers which could sometimes decide to change lanes or could drive on surface streets (assuming they had sufficient map data). Really neat, cutting-edge stuff,” Seltz-Axmacher said. “Investors were impressed. It didn’t matter that that jump from ‘sometimes working’ to statistically reliable was 10–1000x more work.”
He said Starsky believed it didn’t need “true AI” to be a good business; in fact, Seltz-Axmacher believed Starsky could succeed developing the technology and then serving as the operator of the vehicles. Investors disagreed.
It didn’t help that the public stock offerings of WeWork, Uber and Lyft failed to impress and soured investors on technology companies.
“If we showed anything at Starsky, it’s that this is buildable if you religiously focus on getting the person out of the vehicle in limited-use cases. But it will need to be someone else to realize that vision.”Stefan Seltz-Axmacher, Starsky Robotics co-founder
“Unfortunately, when investors cool on a space, they generally cool on the entire space,” Seltz-Axmacher wrote. “We also saw that investors really didn’t like the business model of being the operator, and that our heavy investment into safety didn’t translate for investors.”
Seltz-Axmacher said that if Level 3 autonomy is the equivalent of human drivers, “it’s unlikely any of the current technology will make that jump.”
“Whenever someone says autonomy is 10 years away that’s almost certainly what their thought is,” he wrote.
In his conclusion, Seltz-Axmacher remained optimistic his dream would come true.
“If we showed anything at Starsky, it’s that this is buildable if you religiously focus on getting the person out of the vehicle in limited-use cases. But it will need to be someone else to realize that vision,” he wrote.