Minding the Robot-Minders: Will Autonomous Vehicles Deliver on the Promise of Zero Road Accidents?

Lightly regulating the autonomous vehicle industry evokes a “Wild West” mindset. The clearest demonstration of this occurred on 18th March 2018 when a self-driving Uber hit a pedestrian on Mill Avenue in Tempe, Arizona. On 19th November 2019, the National Transportation Safety Board (NTSB) delivered its findings regarding this fatal accident. The NTSB cited 20 factors that contributed to the accident. Several of those 20 factors contributing to the accident were related to minding the robot-minders.

With respect to the term, ‘robot-minder’, I mean not only the human safety driver, who had the responsibility of being the fail-safe, last safety feature of the self-driving vehicle but also those who had the responsibility of training and minding the robot-minder.

The NTSB and the Tempe Police Department alleged that the human safety driver was inattentive just prior to the accident. Further, the NTSB indicated that those tasked with minding the robot-minder did not routinely check up on how well the robot-minder was minding the robot. The in-car video of the robot-minders was infrequently accessed to check on them. Without this one level of accountability, a robot-minder could assume that his/her in-car activities were unlikely to be monitored.

The training of the robot-minders, although thorough, seemed to neglect the importance of emphasising to the robot-minders that they were the absolute last safety feature in the self-driving vehicles. When combined with a decision to remove co-pilots or co-robot-minders from the self-driving vehicles one may conclude that subtracting redundancy and accountability is a recipe for errors. The kind of errors that are remembered in infamy.

Redundancy and accountability enhance reliability and hence, safety. When governments and the autonomous industry agree on this formula, the promise of zero road accidents isn’t empty, but rather can be fulfilled.