How Will Self-Driving Cars Be Programmed to Deal With Dangerous Situations?

Sat, 6/2/2018 - 7:24 pm by Kirsten Rincon

At the moment, trying to come up with the best way for programming a car to react in the safest and least damaging manner in a situation where a collision is inevitable, is one of the biggest challenges that everyone working on autonomous driving technology is facing.

Who Should Decide How Autonomous Cars Should React?

The decision on what a self-driving car should do when faced with a moral and ethical dilemma can not be made by automakers on their own. To make this decision, experts in several different fields should be included, such as psychologists and philosophers, in addition to automotive engineers and software developers.

Preparing a driverless vehicle for a situation in which it is about to get involved in a collision that can not be avoided, is a very difficult task. For instance, a driverless car is headed towards a group of pedestrians or cyclists, or a vehicle that carries a large group of people, and the only thing that the autonomous vehicle can do to avoid hitting those people is to swerve sharply as soon as possible. But, there are concrete walls on the sides of the road, so it’s highly likely that swerving would result in a serious injury or death to the vehicle’s occupant.

What Is the Least Damaging Solution?

The dilemma is whether to program the vehicle to swerve and risk killing its occupant, in order to avoid colliding with the larger group of people in front of it, so that it makes the least amount of damage and save the highest number of lives possible, or save the car’s occupant, no matter the cost.

It’s difficult to reach a consensus on this question, and it will be a while before a final decision is made on what self-driving cars should do when faced with an ethical dilemma.

Handling Bad Weather Conditions

When it comes to how to program a self-driving car to handle adverse weather conditions, at the moment, the idea that prevails in the automotive world is to program the vehicles to stop automatically when faced with conditions they are not capable of operating in, or allow a vehicle occupant to take control in that particular type of situation. That’s the solution that Google currently uses for its self-driving prototypes.

Long Way to Go Before Consumers Can Purchase Driverless Cars

Finally, when talking about driverless cars, questions such as “When will they finally hit the road?” and “What will they cost?”, are inevitable. Before anyone can start thinking about answering these questions, the above-mentioned dilemmas and concerns will have to be resolved.

When it comes to the number of self-driving cars currently on the road, it’s important to note that all such cars are just prototypes built for testing purposes, and are not available commercially. At the moment, there are several dozens autonomous cars being tested on public roads across the United States.

There are various predictions as to how many driverless cars will be on the road in the future, with most recent estimates putting that number at about 10 million by 2020.

In terms of availability and costs, driverless cars are expected to be significantly more expensive than conventional cars during the first couple of years after their introduction to market, and the average consumer will probably will not be able to afford one. Later on, as the technology continues to advance, they will surely become more affordable and should become mainstream, but it won’t happen for at least another decade.