By on October 30, 2020

Unaware that the inherent danger of motorsport is often what makes it popular (check the ratings for any series throughout history and count the number of driver fatalities if you’re in doubt) Roborace plans on becoming the first global championship for battery-driven autonomous cars programmed to run the course without help. Organizers are convinced that the sport will eventually yield compelling competition with teams using nothing more than their own coding acumen and self-driving hardware. Chassis and powertrains are shared between vehicles, making this a battle of real-time computing algorithms and artificial intelligence technologies.

It actually sounds kind of boring. But one of Roborace’s first live-broadcasted events opened with a bang after one of the cars pitched itself directly into a wall — suggesting organizers could still give the viewing public what it wants.

That hasn’t kept me from finding the sport generally loathsome, however. After piggybacking off Formula E for a couple of years, Roborace was finally ready to upend automotive competition as we know it and herald in “motorsport 2.0,” which basically involved seeing if AVs could reliably go around a racetrack. Season alpha, so named to underline the emphasis on development/programming, introduced DevBot 2.0 and allowed the autonomous racers to be controlled by actual racing drivers. It also helped show that this “sport” was really more of an extended proof of concept — a tradition that has continued through 2020.

Season beta is supposed to foreshadow Roborace’s official championship but it’s clear that organizers are still futzing around with formatting. There’s even a “metaverse” being introduced that incorporates virtual anomalies that cars have to account for. That can include everything from fake walls they’ll have to swerve to coins they’ll want to hit to gain points. Roborace has even said its systems can input digitized weather patterns that cars will have to contend with, which sounds like bullshit since they don’t affect the track in any real way. It honestly feels like a last-ditch effort to make events seem more exciting as viewership has clearly become an issue.


None of this is exciting because none of it’s real and there aren’t any stakes without a human driver. Roborace doesn’t even have full teams present for events due to the pandemic. Practically everything that doesn’t involve physically transporting the vehicles onto the track is being done remotely, making the Acronis SIT that crashed on Wednesday the only incident at Thruxton worth mentioning. It’s hardly the first time an autonomous vehicle suffered a malfunction/miscalculation that ended in a wreck at Roborace. But it gave your author an idea he hopes organizers will embrace.

While Roborace seems a solid way of developing autonomous and AI technology, it’s just awful to watch. There are no drivers to cheer, no manufacturer rivalries to keep track of, and the novelty of self-driving cars is quickly wearing off. But here’s where the sport could turn things around and actually make things engaging for fans. Rather than having these cars go around a track mimicking Formula E, why don’t organizers set up rallycross sprints that are far too dangerous for human drivers to consider tackling? Ramps, moving barricades, spike strips and more could be incorporated — though one would imagine the cars themselves would have to be modified to be more cost-effective if catastrophic damage becomes the rule and not the exception.

CEO Lucas di Grassi has already decided to scale things back and change events around to incorporate opportunities for real drivers to go head-to-head with autonomous cars. Why not allow them to pilot the cars remotely in a joker lap before the computer is required to regain control of the vehicle?

This is likely wishful thinking on my part. Whatever costs are incurred by transforming Roborace into a demolition derby probably won’t be offset by the uptick in viewership. But something clearly needs to change when the only thing anyone seems to care about was an autonomous car that comically pinned itself to a wall before making it down the first straightaway. The sport won’t endure without some tailoring and these promised flashy computer graphics just aren’t going to cut it. Thankfully, Roborace seems relatively self-aware and has repeatedly had leadership suggesting the series would be a groundbreaking addition to motorsport or spectacular failure.

[Image: Roborace]

Get the latest TTAC e-Newsletter!

15 Comments on “Opinion: Roborace Crash Offers Sport Some Much Needed Excitement...”

  • avatar

    Sorry, but I can’t get into the concept. This is attempting to take motoracing and cull it down to the technical differences between the vehicles. Take out the human element removes the challenge, the ability for the human element to compete against fatigue, human error, and the possibility of bad decisions, not just against other vehicles.

    Sounds boring as hell to me. I got over remote racing slot cars around the age of fifteen, 55 years ago.

  • avatar

    Finally, a sport that might threaten the dominance of mighty WNBA women’s basketball.

  • avatar
    Mike Beranek

    Since there’s no people in the cars or on the track, they should spice things up by adding machine guns to the fronts of the cars, like Bond’s A/M. Instead of taking the air off the other car’s spoiler, just blow them away.
    Not that I would watch.

  • avatar

    So with no drivers, what’s the point of a cockpit with windows?

    • 0 avatar

      No need for the raised lump in the middle of the car at all.

      Now, if they raced in fog, snow/dust storms, driving rain, mud, darkness and amongst random, unpredictable, moving obstacles it might be interesting. With the side benefit (if you choose to see it as such) of helping prepare autonomous systems for the real world.

    • 0 avatar

      “So with no drivers, what’s the point of a cockpit with windows?”

      What is the point of the whole car? At least without a driver the cars could get within 6 ft of each other.

      Racing lost something when the cars reached the point of being too fast for the drivers. So in some sense this would be the next logical step except that it wouldn’t be likely to draw much in the way of sponsors — the necessary emotional rivalry won’t work with geeks behind keyboards as the heroes:-)

  • avatar

    for a more realistic scenario there should be a human passenger asleep in the car as it races.

  • avatar

    They should make it like Battlebots, but with racing. Speed vs. firepower!

  • avatar

    So then ;

    Mayhem but no carnage ? .



  • avatar

    Yeah, and this accident happened before the car crossed the start line of the race. It was as the field was accelerating to the start. Boink! Ha ha, some dolt got the detail programming wrong, because they all use the same standardized base programs. Or there was equipment failure.

    Thirty odd years ago, when I raced 1/8 scale gas/nitro R/C cars, AWD, rear torsen differential, front overrun clutches. independent suspension all around you built from a base kit and machined the rest, 3 horsepower at 40,000 rpm and brutal acceleration, when a steering servo went wonky or you had radio problems, well WATCH OUT, spectators! A six pound missile doing over 50 mph will take off your ankles. Or climb a nearby building to rather surprising heights, shattering even carbon fiber chassis’- yeah we had those in the early 1980s (carbon fiber is like the common cold — been around for ever but never really econonomical to use and no one ever cured the common cold), but went back to aluminum. A one hour race taxed the concentration like nothing else I’ve ever experienced. It used to look so easy to spectators, you’d get dolts approaching you telling you they could do better. Right. You’ve spent a dozen years practising, have you, Mac? I used to let them have a go, and they couldn’t even drive straight let alone make the first corner. Three minutes at that and it dawned on them, um, maybe this IS difficult, and they all buggered off tails between their legs. Video games? Who needed that stuff except as minor entertainment? In R/C racing, your thousand dollar investment was on the line if you screwed up and it was the real thing where screwing up had at the very least monetary consequences and lost pride. Virtual reality it was not.

    Tesla is just now releasing Self-Drive to the world, ready or not. So why not put some of those updated woohoo wowee cars in a race with the disengaged driver watching a You Tube video on the big screen and letting the Tesla wundercars navigate the racetrack at speed by themselves? At least then there’d be human lives at stake, and the crowds could watch in deep rapture waiting for the big one. Racing would improve that autonomous car breed immensely, I’m sure. The rats trapped behind the non-steering wheel would insist on it.

  • avatar

    > Unaware that the inherent danger of motorsport is often what makes it popular (check the ratings for any series throughout history and count the number of driver fatalities if you’re in doubt)

    And you posted this on the weekend of the Imola race too. To hell with TTAC, I’ve been here since 2008 and seeing this about of vileness pass the editorial standards is beyond belief. I’m out.

  • avatar


Read all comments

Recent Comments

  • conundrum: Renesas of Japan was THE major automotive chip supplier with a third of the global market until their...
  • Inside Looking Out: Doesn’t Tovares afraid of being cancelled from STellantis for making such outrages...
  • conundrum: Well, the math the EU bureaucracy came up with to prove EVs were so wonderful over a decade ago was a load...
  • DAC17: Why does anyone care about this ridiculous way to strip money out of peoples’ pockets for unnecessary...
  • Inside Looking Out: Did you mean President Yeltsin?

New Car Research

Get a Free Dealer Quote

Who We Are

  • Adam Tonge
  • Bozi Tatarevic
  • Corey Lewis
  • Jo Borras
  • Mark Baruth
  • Ronnie Schreiber