Back in July, 2022, two different deaths of motorcyclists in fatal crashes involved Teslas suspected of using the Autopilot feature. In both cases, the Teslas—which were different models and years—plowed into the riders from behind. Both crashes happened at night, and both involved cruiser-type motorcycles with low-set, close-together taillights.
Why do those details about the bikes matter? In this video, Ryan from FortNine lays out a well-reasoned theory about why Autopilot might not have seen the riders in these cases. To be clear, it’s just a theory—and the National Highway Traffic Safety Administration, as well as other authorities, are still investigating both crashes.
That said, the theory seems pretty solid. See, Teslas used to rely on a combination of multiple cameras and radar to help functionalities like Autopilot calculate safe distance to the vehicles and other objects around a given Tesla. However, Tesla has been phasing out its use of radars across its model range over time, moving them over to its Tesla Vision camera-only system. The most recent such change happened in February, 2022, when Tesla removed radar units from its Model S and Model X, according to our colleagues at InsideEVs.
Not A New Problem:
Now, even if you’re a person who doesn’t really care much about cars, and who mainly focuses your energy on motorbikes, you’ve probably heard a lot about radar-adaptive cruise control in the past few years. Multiple motorcycle OEMs have been rolling out their takes on such systems on their most premium models. Although the systems may work slightly differently, the idea is the same—all rely on radar to help the bikes maintain a safe distance from vehicles around them. Moto OEMs don’t call it “autopilot” for a number of reasons—but it performs some of the same functions, only it relies on the proven accuracy of radar.
Why not use both radar and cameras, you may ask? As Ryan F9 points out, although Tesla (and most companies) will not directly state that it wants to make money as a primary motivational factor, businesses are, after all, businesses. Therefore, the idea that they want to make money should certainly not be news, and should probably be regarded as an implicit factor in all their decision-making. Individuals and companies may also want to do good things, or to be perceived as doing socially-good things by the public—but the underlying motivation to make money is never not a factor.
The placement of those taillights on both bikes involved—a Yamaha V-Star and a Harley-Davidson Sportster—matters if both cars were relying on Tesla Vision to tell them where the next vehicle on the highway was. Tesla Vision relies on a system of multiple cameras and AI, which interprets the visual data gathered by the cameras.
At night, it’s difficult to make out vehicle silhouettes—but you can usually see taillights. If the AI perceived those close-together taillights as being faraway car headlights instead of a much closer motorcycle—well, there’s the problem. Had a radar been involved, perhaps the problem could have been avoided—but we can’t know for sure.
What we do know is that two riders are dead—and also, that NHTSA data released in June, 2022 revealed that “Tesla’s vehicles have been found to shut off the advanced driver-assistance system, Autopilot, around one second before impact,” according to the Washington Post. That observation came from analysis of 273 previous Tesla Autopilot-involved crashes.
What can riders do? Apart from reaching out to your local legislators to demand action, the most immediate suggestion that Ryan F9 has is that if you’re riding on a highway at night, and you see headlights behind you—move from side to side in your lane. Hopefully, that movement will alert the possible AI behind you that you’re a vehicle that’s much closer than it might expect. It’s a frustrating bandaid on a serious problem, but at the same time, it might not be terrible advice to keep in mind.