Various pundits are projecting that by 2020 – just four years – the driving of cars and trucks will be completely automated. Vehicle services, whether for consumers or businesses, will be readily available for very reasonable prices. I will not need to own a personal vehicle and my business can dispense with its fleet of delivery trucks. Taxi or truck drivers will no longer be professions; they will go the ways of elevator operators, bank tellers, and secretaries.
Personal cars are typically in use 5% of their lives. The maturation of autonomous vehicles will cause utilization to approach 90-100%, dramatically reducing the number of vehicles needed to meet demands. The greatly reduced number of vehicles will, in turn, enormously decrease the demand for independent servicing of vehicles – the vehicle service companies will handle this. The number of gas stations, repair shops, and car washes will precipitously drop.
Reduced production of vehicles will result in the supplier base for vehicle manufacturers being steadily weaned due to less demands for nuts, bolts, rims, tires, floor mats, etc. Auto insurance, the cash cow of the insurance industry, will be transformed once accidents cannot happen. People will not pay to avoid risks that are no longer at all likely. I have read projections that 5 million jobs will disappear.
This all seems plausible if the automation performs flawlessly. Then, I can sit in the back seat and snooze on the way to the office, train station, or airport. To evaluate this possibility, I have used my frequent Uber rides as an experiment where I ask myself, “What if there were no driver in the car?” Two examples are very revealing.
I was going to the airport for a flight to Los Angeles. The Uber app guided the driver to take a route that would have taken us away from the airport, not towards it. I told this to the Uber driver and he said, “I have not driven this route before. If you know a better way, just tell me what to do.” Everything worked out fine. What if there had been no driver?
More recently, I was taking Uber to work on a very cold morning. I saw that the driver who would pick me up was just three blocks away and would pick me up in three minutes. As I watched his car on my phone, it did not move and the time increased for 3, to 4, to 5, to 8 minutes. I called him, and he said he would be there in five minutes, which he was.
Upon entering the car, I asked him, “You weren’t in your car when you responded to my request, were you?” He responded, “No I was in my apartment, headed to my car in the parking garage. It took me a few minutes to get there and then drive the three blocks to you.” Driverless cars are unlikely to have apartments, but how will customers deal with departures from “normal” operations?
One extreme possibility is that drivers will be able to ignore vehicle operations until just before an imminent accident when the vehicle signals that the human driver should take over. They will not have manually driven a vehicle in perhaps months and they now have a few seconds to get back in the loop. Not surprisingly, this will not work. Decades of research have repeatedly shown this.
The impacts of autonomous vehicles on the overall economy will be pervasive. The impact on the notion of “driving” will be profound. We need to explore these futures deeply and carefully. Like the canal barges, steamboats, railroads, automobiles, and airplanes, these changes will inevitably happen. How can we better understand and facilitate such changes to assure positive outcomes?