The scene at California's Thunderhill Raceway was decidedly more tame. Located three hours north of San Francisco in the golden foothills of the state's Central Valley, Thunderhill was hot and windy that weekend. Jim Burke, co-founder of the Power Racing Series, and Eli Richter, a member of HackPGH, unloaded a vehicle from the back of their truck by hand. It was, ultimately, a one-man job: the SLAMborghini, as it had been christened, was a matte-black Lamborghini Power Wheels car, outfitted with a set of low-cost sensors to enable basic autonomous capabilities.
The Power Racing Series crew was one of a dozen or so teams that attended Self Racing Cars, an autonomous vehicle race series recently launched by Silicon Valley entrepreneur Joshua Schachter. With no restrictions or qualifications, the event attracted a motley crew of established parts suppliers, startups, investors, and media outlets eager to catch the spectacle. Comma.ai was in attendance, fresh from their press day in Las Vegas, as were autonomous vehicle software developer PolySync, automotive supplier Denso, Chris Anderson's Right Turn Clyde go-kart, and others. Renovo showed up with its perfectly polished $529,000 electric Coupe, but left the self-drifting Delorean, Marty, at home.
Ahead of the weekend, it was unclear what to expect. Outlets like Jalopnik promised the event would "push driverless cars to their limits." Schachter himself was keen to manage expectations. "I'll just have an event and see who shows up," Schachter told The Verge in April. "If people showed up with a self-driving jalopy, I'd be thrilled."
Ultimately, the event was part track day and part marketing demo, with well-moneyed types circling the parking lot and eying engineers. But more than anything, the first installment of Self Racing Cars felt like an engineering hackathon.
In that sense, Schachter said, the series harkens back to the birth of motorsports. "When racing started you didn't just buy a car and start racing it — you had to build a car," he said, standing in front of the wind-swept track. "At its inception, racing was as much about engineering as it was competition."
Burke, of Power Racing, echoed the sentiment. "This event at Thunderhill is incredibly valuable to the autonomous community," he said. "A lot of this technology, you hear it in the news all the time — but there's still a lot of testing miles." Burke and Richter had other motives, too: they're trying to launch a national low-cost ($1,000 and less) autonomous Power Wheels series for high schoolers. "We're here to make the connections to help our series grow," Burke said. "We need someone to help us grant write, or underwrite, or anything. We're as grassroots as it gets! We're punk rock racing."
The track days were broken up into 25-minute slots, with each team granted multiple runs to test the mettle of their machines. Faced with a twisting, no-holds-barred track, the vehicles revealed their shortfalls quickly: PolySync's autonomous braking system malfunctioned, preventing a fully autonomous run; the SLAMborghini could only operate under remote control; and Comma.ai's new $50,000 GPS unit failed — the team was forced to rely on a program their intern wrote on short order.
"All of our technology sounds great on paper," said 3D Robotics' Chris Anderson, sitting in the back of his cargo van. The go-kart he built with Autodesk president and CEO Carl Bass was piloted by a stuffed gorilla, and had suffered a radio link failure. "The beauty of this [track day] is that it's going to find every possible flaw."
That's not to suggest that none of the cars found their way around the course. AutonomouStuff's pearl-white Lincoln MKZ, for instance, slipped around the course with little difficulty and Comma.ai's 2016 Acura ILX completed a fully autonomous lap — almost.The car reflects my inability to drive a racecar," founder George Hotz said after jerking the wheel to avoid spilling into the dirt.
A race, though, may not be the ultimate goal here. "What's the point having amateurs redo what professionals have already done?" Anderson said, pointing out that Google's autonomous car would have no trouble completing a lap better than most of the vehicles present. "I think the point is exactly the same point that Jobs and Wozniak had when they had the Homebrew Computer Club. The Apple II wasn't the first computer or the best computer — in fact it was probably the worst computer — you could buy. But it was the only computer that you could buy, and it took a different evolutionary path."
"The hope here is that this will go in a different direction. That we're not just going to end up with another Google autonomous car. That we're going to do something that Google wouldn't dare, or want, to do."
For Anderson, that means doing something "dangerous, silly, and cheap" — something like a real-world Mario Kart. "I don't know how to make the banana peels big enough," he added with a sigh, "that's my big problem."
Schachter's ambitions were more modest. "I want people to field a competitive vehicle for less than 10 grand," he said, and create an alternative to glossier series like Roborace, set to launch next year. He expressed some reservations about bringing Mario Kart to life, but remained diplomatic.
"I think we need to get Chris to the track more often," he said. "He'll change his mind as to what this could be."
Schachter has experience on the track — he races spec Miatas throughout the year — but other attendees were brand-new to it. Schachter brought in Doug Juenke of Hooked on Driving, an organization that offers driver coaching, to help with safety and organization.
On the second day of the event, I found Juenke in the parking lot, standing over the open hood of his Subaru WRX. Tall, blonde, and tan, Juenke stood out from the crowd.
Juenke has been racing at Thunderhill for 13 years. He'd heard of autonomous cars, but had never been this close to them before. I asked him what he made of Schachter's Race Series.
"Does the thought of a race series sound cool? No," he said, when we sat down to talk in the clubhouse. "The thing about racing is the human. Cars are pretty much very similar, but there's a reason why Sebastian Vettel is Sebastian Vettel." But Juenke recognizes the value of the engineering challenge, and transformational promise of autonomous vehicles. He has no illusions about the future of driving.
"As this younger generation becomes less accustomed to having cars to drive, it's naturally going to become obsolete for the masses. Do I see this" — he nodded to the parking lot outside — "make what we do go away eventually? Unfortunately, I do."
"We're going to drive till we can't drive, until they yank our licenses," he said, gesturing to the driver next to him. "But I think of someone who is six or seven or eight years old now… what are cars going to be like in 10 years?"