In a previous post, I discussed electric vehicles’ (EVs) emerging revolution in transportation propulsion, and how cheaper and cleaner, better-driving cars, fueled by solar roof electricity, should clearly favor suburban development. Today, I want to address autonomous driving (AD) technology, separately from EVs. Certainly, it is easier to apply AD technology to EVs than to traditional internal combustion cars, but EVs are not mandatory for AD, and they certainly provide entertaining and economical driving without it.
Two basic claims about AD make the case for its currently widespread R&D and for its eventual success and adoption:
AD makes auto travel cheaper and more convenient to drivers, returning time to them by assuming driving duties.
In principle, many AD cars driving together create the possibility of a controlled traffic flow—which implies faster, safer travel and a potentially greater vehicle capacity for existing roads.
The first claim is hard to dispute—a “robot chauffeur” should certainly reduce the stress and time costs of travel. But if that’s true, think for a minute about its ramifications: Won’t households be willing to travel longer distances and to own (or employ) multiple robot cars that make more trips to pick up packages, take the kids to school, drop off goods, etc.? Robot car trips will surely increase vehicle miles traveled, and with that comes the potential for greater traffic congestion, partially defeating the purpose of AD.
This additional induced travel demand would offset some of the improved road performance AD might deliver via the efficiencies of vehicle coordination. Moreover, if I could take an AD car from downtown Manhattan directly to my MIT doorstep, driving at higher speeds on a coordinated highway while I sip coffee and work, why would I ever take the train or a plane? Public transportation, which carries many more passengers per vehicle, could suffer a substantial loss in demand. Putting those passengers back onto already crowded roads in individual AD cars could totally offset the promised benefit of better traffic flow through coordination.
To my thinking, suburban highways and roads are where these benefits of AD will have the greatest potential—for two reasons. First, suburban roads and highways are less likely be congested in the first place, offering the best technical circumstances for coordination to actually work. Second, the potential expansion of road capacity is more important to suburbanites, who travel more and spend far more personal time driving. The purported benefits of AD should simply have a bigger impact for suburban households. As city streets are already overwhelmed with traffic, coordinated driving on them is much more difficult to implement.
For a glimpse into these issues, follow the discussion and debate occurring in New York City and San Francisco on whether a much smaller improvement in urban transportation technology—ride sharing à la Uber and Lyft—is beginning to create serious systemic traffic problems.