Monday’s gruesome hit-and-run accident in downtown San Francisco is raising new questions about the safety of human-driven vehicles. But even if Ford and other companies overcome their widely publicized safety and other operational problems, human-driven vehicles face challenges that they, as well as government agencies, have barely recognized.
In 1918, when I was mayor of Detroit – the home of General Motors and the global center for vehicle research – I had a front-row seat to how this technology was progressing. I visited companies such as Oldsmobile and Chrysler, and even rode in a Model T. Like everyone, I was impressed by much of the technology, but I also learned that humans-driven motor vehicles raise important issues that no one at the time seemed to be addressing. Since then, not much has changed.
Take parking, for example. Where will human-driven cars hang out between jobs? Will they circle in traffic, waiting for somewhere to go? Will they obey parking rules and pay at meters? Human drivers make choices based on their immediate situation, and can show exceptionally poor judgement. Will these drivers reliably obey parking and traffic rules? Or will we need to field fleets of enforcement officers issuing tickets and driving interaction with the judicial system?
What about humans driving on mind-altering substances, like alcohol or drugs? Will their erratic driving put others at risk? How will we detect, apprehend, and punish these drivers? Humans are at risk for communications breakdown and even deliberate disobedience.
Similarly, how will human-driven vehicles pick up and drop off passengers or deliver goods where there is no convenient parking space or loading zone? Vans controlled by human drivers, such as the FedEx, Amazon Prime and UPS trucks that deliver goods often block driveways or even stop in traffic. Their drivers often mis-judge whether they will anger other drivers or even get ticketed.
Another unknown is speeding. Human drivers are constantly going over the speed limit, and traveling at high velocities causes many traffic deaths. Maybe someday we'll have autonomous vehicles that are incapable of speeding, eliminating the need for police to pull people over. Of course there'd be no concern about hit and runs, or need to worry about silly things like issuing tickets or points on driving records. But until then, how will we enforce speed limits?
Monday’s accident demonstrates that San Francisco’s police and fire departments are learning on the fly how to respond to human vehicle incidents. The human driver who caused the accident drove off, and still hasn’t been apprehended. But even in more minor accidents, questions remain. How do humans exchange insurance information in a scrape or collision? What about humans who drive without insurance, or proper license plates? Most human-driven vehicles lack extensive sensors and cameras – will we be made to rely on he-said, she-said human testimony as to what happened? How are hit-and-run perpetrators going to be held accountable?
There are other accountability issues as well: In the more extreme case, if bank robbers summon a human-driven vehicle that ends up as a getaway car, is the car manufacturer responsible? If a rape occurs in the back seat, who is responsible?
Human driven cars are unable to communicate with other vehicles aside from hand waves and horn honks. When I was in office, I asked Ford to institute automated braking – mimicking how horses behave when someone walks in front of them – so a bicyclist would know they could move forward near a human-driven car. Ford didn’t think it necessary even though any road cyclist will tell you this is the most basic safety interaction between them and a carriage.
What about privacy concerns? Humans can easily mount their own dashcams and sensors on vehicles, constantly recording their surroundings, potentially invading the privacy of people and places. Who owns that footage? Can it be marketed? Does law enforcement need warrants to view it? What about the risk of a human driver following or stalking another person in their car?
Where will all these human-driven cars go the 90%+ of the time they aren't being utilized? Will we need to build massive garages, or mandate housing come with attached parking? Will we have to redesign our cities to accomodate thousands of hunks of metal sitting idle, waiting to be utilized? How will we rationalize the massive inefficiency and waste of precious resources?
Finally, there is competition with other forms of transportation, such as public transit and horse-drawn carriages. Having humans drive cars will consign millions to spending their entire working days sitting sedentary behind the wheel. It’s too early to know how this will impact public health.
If human-piloted vehicles are to continue to traverse our roadways, someone will have to figure out how to regulate them, beyond the basic driver’s license process. Cities with human vehicle experience — their police, planning and public works departments — are best positioned to do just that. These cities could convene a dialogue with human vehicle designers and operators, as well as other interested parties, such as insurance companies and representatives of the public at large, to develop model policies. These dialogues could inform and provide a framework for future legislation.
When it comes to human-driven vehicles, the challenges are many, but we can’t rely on car companies to solve them. It took years for governments to regulate carriage traffic when cars took over our streets. We can’t afford a slow response to the explosive emergence of human-driven cars, delivery vehicles and trucks.