Choice and Speculation

Abstract

Speculation about how driverless vehicle technology will cities appears just about everywhere. As a transportation scholar, I often am asked to join in, but I have a problem with doing so. According to most speculation, driverless technologies will transform things. Technology is always the actor, like some unalterable force that sets the terms by which cities and human life will unfold. Individuals, governments, and businesses have choices about how they create, sell, and use technology, however, even if that technology promises to be important. We have choices about how we distribute the benefits and burdens wrought by driverless vehicle technology. Those social, economic, and political choices can influence human life in cities just as much as, if not more than, the technology changes, and those choices will shape the technology as much as the technology will inform and influence choice.My speculations in this article, therefore, highlight the places where I think critical changes and choices are likely to appear. I shall focus my comments on two fundamental changes the technology enables: (1) the labor savings that can result from replacing human with machine labor, and (2) the separation of vehicle from owner/operator made possible by the technology.Many people who speculate about the labor savings tend to focus on one transport sector at a time, and by far the most attention has been paid to what machine operation can do for individual drivers by giving them back the time they currently spend operating the vehicle. Also important are the aggregate safety benefits expected from computerized operation in both passenger and freight sectors. The World Health Organization estimates that 1.25 million road fatalities occurred globally in 2013.1 Although it will take some time before the technology becomes global, we should welcome anything that might reduce that horrendous loss of human life.Some individuals object that they do not trust computers to be safer drivers than humans, but that strikes me as vastly overestimating human competencies. Robotic medical and surgical practices and precision manufacturing suggest that machines can have faster, more accurate reactions than human beings do, even if those machines are not perfect. Computers, even if they might be hacked, do not struggle with distraction, self-interest, or thrill seeking. Sensors do break, but they do not get bored, nor do they get drunk or upset. Sensors do not get distracted by a bee in the car, nor do they text their friends when driving. A driverless vehicle turning left at an intersection does not have to guess whether the oncoming vehicle, also driverless, is going to try to squeak through the yellow light; it will know whether that vehicle will stop or go; standard programming should handle that problem. I dream of a future in which I do not have to pound on anybody's car hood (an unfortunately frequent part of my day-to-day life now) because he or she is pulling forward to make a right-hand turn and not looking right for pedestrians like me. Security and hacking are concerns, as always, but fear of potential, unrealized harms seems to me a poor reason to reject technologies that could reduce the existing, unacceptable road fatality and injury numbers we have now.Alleviating drivers of the labor required to drive their own cars has many potential consequences, again depending on personal, business, policy, and planning choices. One major question will be how driverless cars become available to people who currently own their own cars. I tend to envision a subscription service that offers different memberships based on usage, by which individuals can schedule themselves to be picked up and dropped off.2 Urban residents could opt for a lower subscription rate than would subscribers in suburban areas, much as now the former use taxis for occasional trips and the latter tend to own their own vehicles. An engineer in Helsinki, Finland, has published a study to