Should humans be banned from driving? I don’t mean immediately, and I don’t mean to imply they should be replaced with baboon drivers. I mean, when self-driving car technology enables cars to drive themselves with a degree of safety that surpasses humans, should they then become mandatory? It might seem like a weird question, because depending on who you ask, you could get a variety of answers ranging from, “Obviously yes,” to, “Obviously no,” to, “It doesn’t matter, cars are never going to be that good, at least never in our lifetime.” Of course, if you’re following current news on self-driving cars, you know that they very well might be that good in our lifetime. Let’s step back for a moment though, and cover the necessary background information. The most commonly used scale for categorizing the autonomy of cars was developed by the organization SAE International, and rates at a level from zero to five, zero being fully controlled by a human, five being fully autonomous. Level 3 is where you could first call something a “self-driving” car, and it’s where most autonomous cars are right now. Level 3 is automated, but requires a human at the wheel to take over if a situation arises that the car’s systems can’t resolve. Waymo announced in November 2017 that they are testing cars with employees in the backseat, rather than at the wheel(). This means that as of now, level 4 cars are being tested. The Senate Committee on Commerce, Science, and Transportation had a hearing in 2016 with statements from prominent representatives from industry discussing all aspects of self-driving cars to determine how policy should be handled going forward (Hands Off). They raised issues such as federal regulations being necessary to allow autonomous cars to traverse state lines without legal issue, and the need for rigorous cybersecurity on the part of industry leaders. It’s no question that self-driving cars are coming. And when they do, humans should be banned from driving. We’re going to have to overcome the hurdles of trust and perceived loss of freedom, but human drivers should be replaced by self-driving cars for reasons both of efficiency and safety.
If human drivers are able to be outclassed in terms of safety by self-driving cars, they should be replaced, as at that point every driver replaced will help to save lives. It’s not only possible, but likely that self-driving cars will become better than humans. A survey conducted by the NHTSA found that approximately 94% of vehicular crashes were caused by human error (Singh). That 94% is what self-driving cars are capable of reducing. A car doesn’t get drowsy, doesn’t drink, doesn’t undergo road rage, and doesn’t text. It can see in 360 degrees around it and never stops paying attention. A simulation run by Mark Morando et al was conducted to determine the potential safety impact at both a signalised intersection and a roundabout of introducing autonomous vehicles. They found that, “for the signalised intersection, AVs reduce the number of conflicts by 20% to 65% with the AV penetration rates of between 50% and 100% (statistically significant at p < 0.05). For the roundabout, the number of conflicts is reduced by 29% to 64% with the 100% AV penetration rate (statistically significant at p < 0.05)” (Morando et al). While these are simulations, and practical, real world testing is necessary as well, they still demonstrate a substantial improvement. And real world testing is happening, Waymo alone has over 5 million total miles driven by autonomous vehicles (Waymo Reaches 5 Million Self- Driven Miles), and every collision is simulated repeatedly to learn from and prevent a similar incident (Hands Off, 13). This is essential to the safety of autonomous cars versus humans. An accident can be learned from by a company, and a fix can be applied to every car in its fleet. Human error is never going to be done away with so long as humans exist. Safety isn’t the only potential benefit of self-driving cars, though, they have the ability to improve efficiency as well.
Self driving cars can improve throughput of roads, reduce travel times, and improve fuel efficiency, and the effects are more pronounced the more autonomous cars there are. In fact, it’s been estimated that, “cooperative adaptive cruise control (CACC) deployed at 10 percent, 50 percent, and 90 percent market-penetration levels will increase lanes’ effective capacities by around 1 percent, 21 percent and 80 percent, respectively” (Preparing a Nation, 5). This goes to show that it decidedly matters that self-driving cars dominate the road for their benefits to be reaped. Raw throughput isn’t all that will improve, as if self-driving cars really do end up safer, traffic caused by crashes will also go down. The notion of parking will see perhaps the most radical change of all. After all, if a car can drive itself, it can park itself. This means that parking need not be distributed evenly across cities. Roman Zakharenko predicts the rise of “parking belts” in cities, composed of parking structures outside of downtown areas, where cars will park themselves after dropping riders off at their destinations. This leaves more space to be made use of in city centers. There are only benefits to be had from having more self-driving cars on the road, so why isn’t there unchecked public support for them? Well, self-driving cars are a radical change, and people don’t trust them.
Critics point out the many risks and challenges self-driving cars will run into as they develop, from technical malfunctions to the risk of a network of cars being hacked as an act of terrorism, but people are dying right now as a result of driver error, so we have to come to terms with the imperfections inherent to self-driving cars if we’re to cast off the imperfections of manually operated cars. Perhaps the biggest hurdle to mandatory self-driving cars is trust. If people don’t trust them, they won’t allow policy to make them mandatory. An argument someone distrustful of self-driving cars might use is that self-driving cars won’t prevent people’s deaths, there is always a risk of failure in some form or another, so we shouldn’t trust self-driving cars. This line of thinking is the perfect solution fallacy, comparing imperfect self-driving cars to an imagined perfect solution, making self-driving cars look as though they’re failing to accomplish the goal of safety, even if they are able to accomplish better than our current safety. After all, they only have to be better than human drivers to be a worthwhile replacement, so it follows they only have to be better than human drivers for people to trust them more than human drivers. This doesn’t seem to be the case though. A study conducted in 1969 indicates that “the public is willing to accept ‘voluntary’ risks roughly 1000 times greater than ‘involuntary’ risks.” (Starr, 1237). In this case, voluntary and involuntary risks refer to risks that one has a degree of control over versus risks where one has no control respectively. This, then, applies to cars, with driving yourself in a car being a voluntary risk and allowing a self-driving car to drive you being an involuntary risk. This is a massive issue for public trust of self-driving cars. If it holds, then self-driving cars don’t have to accomplish fewer than 30,000 deaths annually, but closer to fewer than 30. All I can ask of people is to be aware of statistics, and be open to trusting technology if data proves we can. Of course, trust isn’t the only thing on people’s minds when asked to take their hands off the wheel.
Self driving cars are not in any way a significant reduction of freedom from driving oneself, and driving should not be considered a freedom worth preserving anyway if it runs such a high risk of fatality. It’s natural to envision the car being taken out of your direct control as a massive loss of freedom. In reality, though, you lose very little when you transfer from a manually operated car to an autonomous one. There already exist rules of the road for people’s safety. Stop lights, lane markers, speed limits, and safety requirements for cars already prohibit absolute freedom while driving. Public roads and the laws that govern them don’t exist primarily for an enjoyable driving experience, they exist to safely and efficiently deliver people to their destinations. Even if you find it unbearable to imagine never driving again, private roads and race courses will still exist for recreational driving, no matter how much market penetration self-driving cars achieve. Finally, consider the legal perspective. The supreme court has held that free speech, even considering its status as a constitutional right, does not extend to creating a clear and present danger, famously providing the example of “shouting fire in a crowded theater” (Schenck v. United States). In driving, while there is no intent to harm, the act of getting behind the wheel naturally puts those around you at risk. The principle of freedom of oneself vs. life and limb of others still holds. I understand the potential joys of driving. America has a strong car culture, and people can feel liberated behind the wheel. But those feelings can’t take precedence over lives.
In volume 100, issue 1 of Scientific American an article was published that said, “That the automobile has practically reached the limit of its development is suggested by the fact that during the past year no improvements of a radical nature have been introduced.” That was in 1909. It seems like science fiction, saying that humans might not be a necessary part of the equation in a car, but it’s possible. We’ve got to be open to a future where it’s true. People’s lives depend on it.
“Automobile and Motor Boat.” Scientific American, vol. 100, no. 1, 2 Jan. 1909, archive.org/details/scientific-american-1909-01-02 p.6.
Starr, Chauncy. “Social Benefit versus Technological Risk.” Science, vol. 165, no. 3899, 1969, pp. 1232–1238. JSTOR, JSTOR, www.jstor.org/stable/1727970.
Morando, Mark, et al. “Studying the Safety Impact of Autonomous Vehicles Using Simulation-Based Surrogate Safety Measures,” Journal of Advanced Transportation, vol. 2018, Article ID 6135183, 11 pages, 2018.
“Preparing a Nation for Autonomous Vehicles Opportunities, Barriers and Policy Recommendations.” Eno Center for Transportation, Oct. 2013, http://www.enotrans.org/etl-material/preparing-a-nation-for-autonomous-vehicles-opportunities-barriers-and-policy-recommendations/.
Schenck v. United States, 249 U.S. 47, 49-51, 1917.
“Taxonomy and Definitions for Terms Related to Driving Automation Systems for On-Road Motor Vehicles J3016_201609.” SAE International, SAE International, 16 Jan. 2014, http://www.sae.org/standards/content/j3016_201609/.
United States, Congress, Cong., Committee on Commerce, Science, and Transportation. “Hands off: The Future of Self-Driving Cars.” Hands off: The Future of Self-Driving Cars, U.S. Government Publishing Office, 2016. 114th Congress, 2nd session, report.
United States, Congress, Singh, Santokh. “Critical Reasons for Crashes Investigated in the National Motor Vehicle Crash Causation Survey.” Critical Reasons for Crashes Investigated in the National Motor Vehicle Crash Causation Survey, National Center for Statistics and Analysis, 2015.
“Waymo Reaches 5 Million Self-Driven Miles.” Medium, 28 Feb. 2018, medium.com/waymo/waymo-reaches-5-million-self-driven-miles-61fba590fafe.
“Waymo’s Fully Self-Driving Vehicles Are Here.” Medium, 7 Nov. 2017, medium.com/waymo/with-waymo-in-the-drivers-seat-fully-self-driving-vehicles-can-transform-the-way-we-get-around-75e9622e829a.
Zakharenko, Roman. “Self-Driving Cars Will Change Cities.” ScienceDirect, 15 Sept. 2016, www-sciencedirect-com.jpllnet.sfsu.edu/science/article/pii/S016604621630182X.