As anyone who has ever tried Tesla's Autopilot or Cadillac's Super Cruise can attest, it's hard to be a part-time driver. You chill out in passenger mode until, occasionally, the car asks you to take over the wheel when, say, there's too much rain in the road or the lane dividing lines are missing and it can't “see.” Humans are inclined to tune out in these sit-back-and-relax situations—an occurrence known as vigilance decrement—and that appears to be one of the reasons why a self-driving Uber SUV fatally hit and killed a pedestrian in Arizona last week. For sure, the LiDAR, radar and camera technology somehow dropped the ball in this tragic accident, but so did the human safety driver, which video reveals wasn't looking at the road in the seconds before the crash. Considering how Teslas and Cadillacs have built-in sensors and cameras to detect when drivers have taken their hands off the wheel or averted their gaze from the road, it's surprising that those systems don't appear to be similarly in place in self-driving test cars. Humans monitoring computers need computer-monitoring, too, or we'll never get to the statistically-safer self-driving car future where drivers truly can take their eyes off (and pedestrians can safely walk on) the road.
sign up for our newsletter