There are more than 5 million automobile accidents every year in the United States, resulting in over 40,000 deaths. Worldwide, there are over 1 million deaths. Because over 90% of accidents are caused by human error, many companies have started investing in research about and development of self-driving cars, with the goal of removing that human error from the equation. While elements of self-driving cars such as parallel parking and lane change assist are already available, people are still waiting for a widely available, fully self-driving car. A fully self-driving car would be able to sense its rapidly changing environment, allowing it to drive with little to no human input. In this article we’ll explore the feasibility of self-driving cars, considering the advancements that have been made and the obstacles that lie ahead.

The Technological Side

Enlarge

SelfDrivingCar2
A Google Self-driving Car

When it comes to technology, self-driving cars have made great strides. Among the many companies developing self-driving cars, Google has been one of the most prominent. Since 2009, Google has been testing a fleet of 32 vehicles. Today, these vehicles log about 10,000 miles every week and have driven over 1.8 million miles since testing began. The technology in these cars is able to detect a car’s surroundings and make adjustments to the way the cars are driving. The car is even able to detect the types of cars that are driving near it, coming to a stop, for example, if it detects an ambulance passing by. Google hopes to roll out commercially available self-driving cars in four years. However, the technology behind self-driving cars is not accident-free. Google’s fleet has been involved in a total of 16 accidents since 2009, all of which were considered “minor.” Google says that each one of these was due to human error. Additionally, self-driving cars may make erratic moves themselves, swerving suddenly if they feel a poorly parked car is attempting to pull into traffic when it really is not. Therefore, the challenge that remains for developers of self-driving cars is to account for irrational and erroneous human behavior on the roads.

The Legal Side

Self-driving cars also come with a wealth of legal issues. Most of the regulation on self-driving cars has been coming on a state-by-state basis. California, Nevada, Florida, and Michigan already have regulation on self-driving cars, but the rest of America’s states do not. Soon, those states or the federal government will have to determine their policies. In 2013, the National Highway Traffic Safety Administration released a set of guidelines for regulation, but the federal government has yet to tackle the issue. Additionally, current driving laws will have to be updated to account for self-driving cars. For example, in New York a driver is required to keep at least one hand on the wheel at all times, but that would become pointless if the car can drive itself. Finally, self-driving cars have legal implications for insurance companies as well, for they will have to reconsider how much to pay in cases of accidents and whether to pay at all. All of these issues will have to be dealt with by the government in the coming years.

The Ethical Side

If a self-driving car is going to run into ten pedestrians, what should it do? Should it keep going and kill the pedestrians, or should it swerve out of the way and kill the driver?

Finally, the concept of self-driving cars raises some major ethical questions, mostly regarding what to do in the case of an inevitable accident with a pedestrian. If a self-driving car is going to run into ten pedestrians, what should it do? Should it keep going and kill the pedestrians, or should it swerve out of the way and kill the driver? As explained by Jean-Francois Bonnefon of the Toulouse School of Economics, this is a major ethical dilemma that will have to be resolved. In most polls taken, people were OK with self-driving cars killing the driver to save a greater number of pedestrians, but would not be willing to follow that paradigm if they were the driver. The issue could also turn into a practical problem for self-driving cars and their sales. If self-driving cars are programmed to kill the driver instead of ten pedestrians because that minimizes loss of life, then fewer people are likely to buy self-driving cars. That would mean, however, that the safety benefits that generally arise with self-driving cars are less likely to be realized. The question also changes when the number of pedestrians changes — what if it were five pedestrians, or just one? In the coming years, this is an issue that will be shaped by developers of self-driving cars, philosophers, and the general public alike.

All in all, self-driving cars have made numerous strides in the last decade, but much work remains to be done. While self-driving cars do have the potential to again change how humans move, developers must ensure that those changes come in a safe and technologically, legally, and ethically sound manner.

About The Author