While autonomous vehicles, powered by sophisticated sensors and cameras, can reliably see their surroundings, the software doesn’t always understand what it detects.

Share story

The tragedy of the first pedestrian killed by an autonomous vehicle points to a potential vulnerability with the nascent technology being tested on the open roads: While robo-cars, powered by sophisticated sensors and cameras, can reliably see their surroundings, the software doesn’t always understand what it detects.

Neither the self-driving system in the Uber  autonomous vehicle that struck and killed a woman in Tempe, Arizona, nor the human safety driver behind the wheel hit the brakes when she apparently stepped off a median and onto the roadway Sunday night, the Tempe police chief told the San Francisco Chronicle. The human driver told police he didn’t see the pedestrian coming, and the autonomous system behaved as if it hadn’t, either.

Experts say that the sophisticated sensors on the autonomous vehicle almost certainly detected the woman pushing her bicycle laden with bags along the median, close to the road. But it’s possible the car’s lidar and radar sensors, which scan the surroundings for objects, may not have realized it was detecting a person. (Uber did not immediately respond to a request for comment.)

“The real challenge is you need to distinguish the difference between people and cars and bushes and paper bags and anything else that could be out in the road environment,” said Matthew Johnson-Roberson, an engineering professor at the University of Michigan who works with Ford Motor on autonomous-vehicle research. “The detection algorithms may have failed to detect the person or distinguish her from a bush.”

Driverless cars “see” the world around them using data from cameras as well as radar and lidar sensors that bounce laser light off objects to assess shape and location. High-speed processors crunch the data to provide a 360-degree view of lanes, traffic, pedestrians, signs, stoplights and anything else in the vehicle’s path. That’s supposed to enable the vehicle to know, in real time, where to go and when to stop. But pedestrian identification remains a major challenge for self-driving systems.

Autonomous vehicles also struggle to master the elements. Snow, ice and rain can obscure sensors and render the most advanced computing power useless. That’s one reason most self-driving cars are being tested in sunny climates like Arizona and Texas.

In the case of the fatal accident in Tempe, the dark conditions aren’t likely to have played a role. While darkness can limit the vision of the cameras, radar functions equally well during the day or at night. Lidar functions better in the dark because the glare of sunshine can sometimes create interference, said Raj Rajkumar, a professor of electrical and computer engineering at Carnegie Mellon University who works on autonomous vehicles.

To Rajkumar, there’s little doubt “lidar would certainly have detected an obstacle.” Any shortcoming would likely be due to classification software “because it was an interesting combination of bicycle, bags and a pedestrian standing stationary on the median,” he said. Had the software recognized a pedestrian standing close to the road, he added, “It would have at least slammed on the brakes.”

Rajkumar said many of the companies he works with require human safety drivers to take the wheel when pedestrians are present along the roadway. “Just as a precautionary measure,” he said. “Nobody wants to deal with this kind of outcome.”

Johnson-Roberson notes a shortcoming in robot reasoning. Had the sensors on the Uber vehicle recognized that a person was on the median, it could have slowed down or hit the brakes as a precaution, as humans do every day when they detect a risk along the road.

“I live in Ann Arbor, a college town,” Johnson-Roberson said. “So on football weekends, when there’s a bunch of drunk college kids, I drive at a lower speed. Those are the kind of human decisions we make to anticipate a situation, and that’s hard with autonomous cars. We’re not there yet.”

The race to get robot rides onto the road has pushed some companies to take too many risks, said Jake Fisher, director of auto testing at Consumer Reports. “There are some out there who are trying very hard to make self-driving cars a reality before the technology is developed,” Fisher said. “A tragedy like this could undermine people’s faith in this technology and actually set the development back many years.”