Discussion Case: The Arrival of Autonomous CarsBright Future or Looming Threat?

As Elaine Herzberg walked her bicycle across a six-lane road in Tempe, Arizona, around 10 o'clock at night, she was fatally struck by a Volvo SUV, a prototype autonomous vehicle that Uber had modified to test its self-driving technology. A series of safety failures had combined to cause this fatal accident. First, the sensors mounted on the car failed to spot Herzberg in sufficient time to slow the vehicle down from its 38 miles-per-hour speed. Second, according to police video cameras, the safety driver, a human sitting in the vehicle charged with monitoring the driving, did not have his hands on the steering wheel and was apparently distracted. Third, the car's brakes were never applied by the safety driver or the car's system. In another incident, a Tesla Model X SUV traveling on Highway 101 in Mountain View, California, slammed into a concrete highway lane divider and burst into flames, killing the driver, Wei Huang. Tesla reported that the autopilot system was engaged and provided warnings to Huang of a potential collision, but Huang failed to take control of the vehicle. According to the company's press release, Huang, "had received several visual and one audible hands-on warnings earlier in the drive and the driver's hands were not detected on the wheel for six seconds prior to the collision. The driver had about five seconds and 150 meters of unobstructed view of the concrete divider . . . but the vehicle logs show that no action was taken." While the self-driving technology was state-of-the-art in both vehicles at the time of the fatal accidents, it had not been sufficient in either case to prevent tragedy. Some thought that the problem was that the humans in the cars had simply been too slow to react to an unexpected event when they thought the car was in charge. According to Constantine Samaras, a university engineering professor, "When humans are the backup systems, we're pretty bad at doing [what is necessary to prevent crashes]. This is a challenge for this transition to automation, where there's this muddled mixture of human responsibility and robot responsibility." Many saw autonomous vehicles as a way to improvenot degraderoad safety. In 2017, more than 37,000 people died in car crashes in the United States. Clearly, humans were not perfect drivers. Self-driving cars, by contrast, did not get tired, frustrated, distracted, or drunk, as humans often did when behind the wheel. A study by the Virginia Tech Transportation Institute analyzed more than 50 self-driving vehicles commissioned by Google, which had traveled about 1.3 million miles on roads in California and Texas. The Google fleet was involved in just 17 crashes over six years, with none of the incidents being the fault of the self-driving vehicles. Another study by McKinsey estimated that autonomous vehicles could reduce U.S. auto accidents by 90 percent. However, others saw negative consequences in the world of autonomous driving. Driverless vehicles could cost truckers, taxi drivers, and other driving professionals their jobs. Already, driverless trucks were being used to move iron ore at mines in Australia, and Canadian-based Suncor Energy had automated its truck fleet, eliminating the jobs of 800 drivers. Companies engaged in other forms of transportation, such as buses, trains, and airplanes, could see fewer customers, as people might choose to travel by car if they were not burdened with driving. If the number of accidents declined, auto repair shops and mechanics would likely have less business, and layoffs could occur. Another issue that critics of autonomous automobiles raised was the possibility that the vehicle's computer systems might be hacked. As part of a planned experiment, hackers Charlie Miller and Chris Valasek could access the controls of a self-driving Jeep Cherokee remotely, instructing it to roll out of a parking lot and into a grassy ditch. The person in the driver's seat at the time, a journalist from Wired magazine, was not controlling the steering wheel or the pedals. Effectively safeguarding against such hacking can be challenging, since self-driving cars have far higher levels of connectivity than human-driven vehicles currently on the road. Will autonomous vehicles be widely adopted in our society? The answer to this question is unclear. In a survey reported in The Washington Post, most Americans said they thought autonomous cars would be quite common within 15 years. Seventy-four percent, however, said they did not expect to own one, and more than 60 percent said they would not want to walk or ride a bicycle anywhere near one. A study conducted by the American Automobile Association (AAA) shortly after the Uber and Tesla accidents reported that 73 percent of Americans said they would not ride in an autonomous vehicle. Greg Bannon, director of engineering at the AAA said, "Any incident involving an autonomous vehicle is likely to shake consumer trust, which is a critical component to the widespread acceptance of autonomous vehicles." Yet, as autonomous vehicle technology gained momentum, governments stepped in to regulate. In 2017, the National Highway and Transportation Safety Administration (NHTSA) released new federal guidelines for Automated Driving Systems (ADS). By 2017, 29 states had enacted legislation governing autonomous vehicles. Some states, such as Florida and Arizona, encouraged the safe development, testing, and operation of self-driving vehicles on their public roads, seeing this as an opportunity to encourage business development. Delaware established the Advisory Council on Connected and Autonomous Vehicles, tasked with developing recommendations for innovative tools and strategies that could be used to prepare Delaware's transportation network for these vehiclesan action also taken by numerous other states. While most people expect that fatal crashes involving autonomous vehicles will occur again in the future, the question of whether self-driving cars impose an acceptable level of risk relative to their benefits to society is a matter of ongoing debate.

Discussion Questions

4. Do you think federal or state governments should regulate the emergence of autonomous driving technology, and if so how?

5. If you were the chief technology officer of a company that was developing autonomously driven vehicles, what steps would you take to manage this technology.