Skip to content

Some Surprising Possible Consequences of Autonomous Vehicles

Share this Story on Facebook, X, Text, LinkedIn, Gmail, Yahoo Mail, or Outlook

My in-box these days is flooded by commentaries and studies dealing with our autonomous vehicles (AV) future. It’s hard to keep up, but three recent items struck me as particularly interesting.
The first was from Scientific American, “How Pedestrians Will Defeat Autonomous Vehicles,” by Karinna Hurley. She summarizes a recent paper in the Journal of Planning Education and Research by Adam Millard-Ball, on the subject of the interaction between AVs and pedestrians in cities. Millard-Ball begins with the reasonable premise that AVs “will almost surely be programmed to avoid hitting people.” But, as a consequence, pedestrians will come to understand this, and the result will be a new cross-walk game of chicken. He then offers three possible outcomes of this new situation:

  • Pedestrian supremacy—if those on foot always win out, a consequence would be much slower car travel in the city, with walking becoming more efficient than motoring.
  • Regulatory response—in which the blame for pedestrian-car accidents is shifted to law-breaking pedestrians, which could be politically untenable.
  • Human driver return—in which many people revert to driving their own vehicles to retain the advantage of faster travel in the city.

My second discovery was a provocative report from Ptolemus Consulting Group called Autonomous Vehicle Global Study 2017. I only had time to read the free abstract, which itself is over 100 pages long. Here are a few of its speculations.
First, as I’ve written in previous newsletters, driverless vehicles could seriously disrupt public transit. Without the cost of a driver, robo-taxis and shuttles could operate door-to-door services where public transport is uneconomic. And, “In turn, this will allow more people to move away from densely populated areas,” a point also made in my third discovery, “Will Self-Driving Cars Make the Suburbs Great Again?” by Nicole Kobie in New Statesman (Dec. 19, 2016). She reports on a presentation by Karen Harris of Bain & Company, based on the idea of spatial economics—as the cost of distance goes down, people won’t mind traveling longer distances, as long as they can make good use of the time. And this could accelerate a trend of professionals moving to exurbs.
Another speculation from the Ptolemus report is that AVs may increase urban traffic congestion. The report cites a study from the Center for Transport Studies, Imperial College London, in which the impact of making car travel as comfortable as light rail or high-speed rail, in terms of smooth acceleration, was weighed against congestion at traffic lights. At least with a mixed fleet of 25% AVs, their scenarios all showed that slower acceleration of the AVs (for a smoother ride) would increase congestion in the vicinity of the intersection—in some scenarios by 50%. In addition, the report looks at what may happen if most AVs are personally owned (rather than being in shared-vehicle fleets). Since the vehicle would go elsewhere after delivering a commuter to the office, its return journey would be empty, significantly adding to the number of vehicles on the road. They also speculate that visions of most people sharing rides in robo-taxis (which means sharing with strangers in a small, enclosed space, with no driver) are completely untested and might be found unacceptable by many people.
These are still very early days in the coming AV world. And regardless of how rapidly the technology improves, there are large unknowns in what choices people and institutions will make, and what the transition to this world will be like. My best advice at this juncture: do not count on any glittering scenario to be anything more than a set of guesses.
(This article first ran in the April issue of Surface Transportation Innovations)
Email this author

Share this Story on Facebook, X, Text, LinkedIn, Gmail, Yahoo Mail, or Outlook

Join Our Email List

Sign me up for:
This field is for validation purposes and should be left unchanged.