After a Tesla Model S operating in self driving mode rear-ended a Culver City fire truck on January 22nd, some major concerns were brought up around the implications of self driving cars sharing the roads with driver-operated counterparts. While the January crash was not the only one involving alleged self-driving cars, the more crashes occur under these circumstances, the more concern there is around self driving cars. So what does it all mean? What are the concerns? Although not a comprehensive list, below are some of the concerns we have here at Project DRIVE.

Communication with Other Vehicles/Sensor Accuracy

One of the biggest concerns around self driving cars is their ability to communicate with other vehicles on the road. In a perfect world, every autonomous vehicle would be able to send and receive information between other vehicles, particularly location information, so the computers can make decisions as to when its safe to merge, etc. However, this would require either every outdated car to be removed from the road or be retrofitted with the necessary equipment. Neither option is cost effective nor practical.

Vehicles with autonomous features currently use a host of sensors to determine other vehicle positions, lanes, speed limits, and other necessary parameters. While sensor technology and accuracy is rapidly improving, there will be situations that no sensor can cope with, such as snow. For example, if a speed limit sign is obstructed by snow, sensors that try to read written text will likely not be able to recognize the speed limit.

When Can the Human Take Over?

Autonomous driving features are in no way advertised as perfect. In fact, many auto maker that offer the feature tell you explicitly that you must be ready to take over at a moment’s notice, but this presents a problem in itself. When a driver is in full control of a car, the driver can execute a corrective action at a moment’s notice. The driver only needs to recognize the need for corrective action. When a car is in self driving mode, not only does the driver need to recognize a corrective action is needed, but that the car won’t make the corrective action on its own. A driver might think “what if I correct the wheel at the same time the car does, which would result in a correction twice as sever as needed?”. Having a machine operate on its own will unavoidably result in an extra step being required before a human can take over, and in term of driving, it could have dire consequences.

Who is Liable When Cars Crash?

When a Tesla in partial self-driving mode killed its driver by swerving underneath a turning semi truck, the NTSB discovered that the driver had been warned multiple times to put his hands back on the wheel, but is that enough? Statistically speaking, as more autonomous cars enter the road, accidents involving them will increase. These can be accidents where a human-operated car slams into a self driving car by mistake as well as mistakes made by the autonomous cars themselves. The biggest question is which party should be liable.

Some will argue that robots and machines should be held liable in any accident. They can take in information and react much faster than humans. However, the current technology does not allow for perfect interpretation of information and still require some human input. If a driver fails to take the wheel after being warned, is the car completely free of responsibility? Additionally, if a car is capable of recognizing a scenario where it needs driver input (i.e. potential danger) shouldn’t it be able to react on its own?

The Bottom Line

There is not clear answer for any of the above concerns. Many of the questions involve ethics and theory, which is often entirely subjective. Technology certainly isn’t advanced enough to create perfect self driving cars. There is also concern over public opinion about self driving cars. If most people hate self driving cars, should they be forced to drive alongside them? Is there a future where every car is autonomous and can communicate fully with each other? A future where these conditions exist is so far away that it would be impossible for us to make a decisions about the topic now. If fact, many of the people alive today won’t be alive to see a future where the previously mentioned scenarios exist. The importance of asking these questions now is important for how we move forward in developing the technology.

What do you guys think about self driving cars? Please comment below whether you guys think they’re a good idea or a waste of time.