What’s Missing from Uber (and Google’s) Push for Autonomous Cars

Uber’s (and Google’s) push for autonomous cars is predictable. We humans love having machines do our work for us. Elevators don’t need operators anymore so why do cars need drivers? Quite simply, all of the safety and technological challenges of transporting people in elevators have been wholly met so now millions of people safely travel in these precusers to autonomous cars.

The differences between elevators and cars are many, of course. First is the fact that an elevator travels within a system where the space of the car’s travel is never shared and always defined. Also, as I mention above, there were decades of elevator use where the elevator car required an operator to control and monitor the car’s operation. So, the evolution went like this: invention of the elevator, operation of the elevator by a human and finally operation of the elevator by a electro-mechanical system (with some human input…the selection of floors, opening and closing of doors and emergency button).

Aviation has seen huge benefits from autonomous systems. Commercial airliners spend a large percentage of flight-time piloting themselves, but pilots have to have a deep knowledge of the complicated systems in the cockpit in order to intervene when necessary. “One of the myths about automation is that as the level increases, you need less human expertise,” said Bryan Reimer, a research scientist at the Massachusetts Institute of Technology and associate director of the New England University Transportation Center. He points to Chesley Sullenberger’s 2009 landing of his US Airway Flight 1549 in the Hudson River. “Automation never would have done that,” at the time, Reimer said. “It doesn’t work outside programming bounds well.”

Cars have been in continuous development for well over a century. ABS brakes, automatic transmissions, traction control, Xenon headlights and a host of other advancements have all added to the overall convenience and safety afforded by the cars that are driven today. Many of today’s cars can alert the driver of obstacles in the road and some can even apply the brakes if the driver fails to do so. Cars can also alert the driver that his car has left its lane or that another car is closer than is safe. What all of those developments have in common is that they are designed to aid and augment the attention, skill and judgment of the driver.

When the Uber car in Arizona smashed into the woman who was walking her bike across the road the system in place was specifically designed to function without the attention, skill and judgment of the driver, and therein lies the problem with the company mindset at Uber.

Uber and the other companies who seek autonomous cars are missing crucial safety steps. Their focus on their ambitious long-term goal has caused them to avoid confronting the myriad of short-term details that must be identified and addressed before the goal can be achieved. In the Arizona case, there was a driver who had so much unwarranted confidence in a system about which she surely had very little actual understanding that it caused her to effectively turn off her own attention, skill and judgment. The driver suffered from a belief (that the car would operate safely by itself) rather than exercising an opinion (that perhaps the car was going too fast for such a dark road) irrespective of the posted speed limit.

That particular brand of technology, the kind that invites us to pay less attention to what we are doing than we would if we had to manage more of the processes for ourselves, is a sign of how foolish and shortsighted big companies can be. If one goal of autonomous cars is protecting human life, then human judgment and decision making must always be valued and encouraged.


What’s Missing from Uber (and Google’s) Push for Autonomous Cars