Are Fully Self-Driving Vehicles A Good Idea?
When most people see self-driving cars in the news, or on social media, they automatically think only of the main positive, the fact that they will no longer need to concentrate on driving. Only thinking of how autonomous vehicles will make driving more straightforward is the opposite of what everyone should be doing, we should be looking at the whole picture and trying to understand precisely how they will affect us in both a positive and negative way. I hope that this story will help provoke thought and interesting discussions on the topic.
Will we be safer on the roads?
There are quite a few benefits to fully autonomous vehicles, one of the main ones being safety. Currently, the leading cause of road accidents is human error, take the human out of the equation, and you will, in theory, have far fewer accidents.
When we drive, we are continually relying on other drivers to obey the laws, drive safely, and accurately indicate where they are going. Relying on other drivers is hard work, especially when many drivers choose to set their own speed limits and seem to have not figured out how to use the indicators. Autonomous vehicles solve this problem, yes the cars are still relying on other vehicles to provide accurate information, but it is far less likely that an autonomous vehicle is going to provide false information.
Self-driving cars also don’t suffer from distraction, they are always on the lookout for possible hazards. Cars always being aware of their surroundings dramatically reduces the risk of colliding with, for example, pedestrians. I say dramatically reduces and not removes the risk because humans are very unpredictable. It is simply not possible for a car to know what every pedestrian is thinking. For example, if a pedestrian suddenly decides to cross the road without looking, there is not a lot the car can do other than react as quickly as possible, which for an autonomous vehicle is a much quicker reaction than a human driver.
Who lives and who dies?
This is an age-old question for autonomous vehicles, in the event of an accident who does the car save, its passengers or the other party. The morality question is impossible to answer definitively.
This question has actually been widely discussed since 1967 and is known as the Trolly Problem. The Trolley Problem asks this:
You see a runaway trolley moving toward five tied-up (or otherwise incapacitated) people lying on the tracks. You are standing next to a lever that controls a switch. If you pull the lever, the trolley will be redirected onto a side track, and the five people on the main track will be saved. However, there is a single person lying on the side track. You have two options:
Do nothing and allow the trolley to kill the five people on the main track.
Pull the lever, diverting the trolley onto the side track where it will kill one person.
Which is the more ethical option?
What would you do in this situation? Save the five people or save a single person? Without knowing anything about the people, you would probably choose to save the five people as the lives of five people are greater than the life of one. But are the lives of five people actually greater than the life of one? You can’t put a price on life, so how can you decide that saving more lives is better than saving a single one.
The Trolley Problem is, of course, a theoretical dilemma, but autonomous vehicles have to come up with an answer to a very similar and genuine problem. The problem itself leads to more questions, who do we trust to write the code that makes this decision? And do we even want another entity to make this decision on our behalf?
When it comes to writing the code, we have to trust the companies developing the vehicles because there is no other option, our lives are in their hands, so to speak. When it comes to who is making the decision, personally I would rather make the decision myself. Drivers tend to instinctively swerve to avoid accidents, even if it is detrimental to their safety, is this what autonomous vehicles should also do?
MIT has developed an experiment called the Moral Machine. In the experiment, you are shown various scenarios, and you have to decide what the self-driving vehicle should do. After a couple of scenarios, you will quickly understand how immensely tricky this problem is. I highly recommend going to the website and judge for yourself how you think the scenarios should play out. I also recommend reading more into the results of MIT’s research so far and the Trolley Experiment. I think it is essential for everyone to understand the problem to the best of their abilities as a world full of autonomous vehicles comes ever closer.
Should there be any driver input?
In short, yes, I think there should be driver input. In my opinion, autonomous vehicles should only be partly self-driving. By this, I mean that they should be able to drive manually with some autonomous features or in a fully autonomous mode. I have a few reasons why I think this, but I also know this causes some problems.
My main reason is that there are a lot of people in the world that really enjoy driving. People who will specifically go out to drive with no destination in mind to relax and enjoy life. This passion for driving fuels numerous sports, ones which would be forever damaged if human-driven cars were banned.
The main issue with this is how do we manage roads where there is a mixture of automation levels. My solution to this is different zones of automation, let’s say three different zones. One zone where a vehicle is required to be fully autonomous, one where it only needs to partially autonomous and one where no automation is necessary.
Fully autonomous zones could be areas like cities, large towns, and motorways where there are a lot of vehicles and a higher risk of accidents. Partial autonomy could be less dense areas such as suburbs and smaller towns where there is a medium risk of accidents. Lastly, no automation could be required in villages and the countryside, where the risk of a collision is significantly reduced. Now, even though the risk of an accident in rural areas is lower due to the lower volume of vehicles on the road, the risk of injury is much higher. The injury risk is higher because cars in rural areas tend to travel at higher speeds than in more built-up areas. This increased risk is why I said that I think vehicles should be able to drive manually but with some autonomous features. The cars should still have autonomous safety features where it takes over control if a medium to high-risk hazard is detected. Say you are driving along a single track country lane, the vehicles should still be communicating with other vehicles that are within range so when you come across a blind corner, it knows if something is coming the other way. Because it knows if something is coming it could then work out if the driver is safely reacting to the blind corner by slowing right down. If the driver is not reacting safely then the car should take over control to avoid a deadly head-on collision.
We are a long way off multiple autonomous zones being a possibility. For this to work, all cars would need some level of automation and progress in the development of autonomous technology is required. In an ideal world, to me, the zones and partly autonomous vehicles are the best way to keep the majority of drivers happy, while still drastically improving safety.
Will our cars get hacked?
Yes, autonomous vehicles will get hacked. Anything with a computer chip can be hacked, and there will always be people in this world that want to do this for malicious reasons. The real question here is how do we prevent someone from taking full control of the vehicle so that we don’t end up in a situation like in Fast and Furious 8. I am not a security expert, nor I am very knowledgable about hacking, so I don’t know what the possible solutions to the hacking problem are.
Is traffic going to be a thing of the past?
This is a difficult question to answer because there are many factors in traffic. One factor is purely the number of cars that are on the road. Autonomous vehicles are unlikely to reduce the overall number of vehicles on the road. There are some factors that self-driving cars could solve, let’s start with one of the major causes of traffic, unnecessary braking and lane changes.
How many times have you been driving down a motorway when someone randomly brakes or switches lanes with no gain? I see this regularly. At the time the slight braking may seem harmless, but it causes a domino effect, a wave of breaking that can stretch back miles and cause congestion. Autonomous vehicles wouldn’t do this, in theory, they would only break when they need to, thus never starting a domino effect. Another cause of traffic is accidents, and this goes back to the safety aspect. Autonomous cars should result in far fewer collisions, resulting in fewer lane closures due to damaged vehicles.
Lastly, we have slow reactions. Humans are pretty slow at reacting to things, such as traffic lights. Autonomous vehicles should be much quicker at responding to a green light turning on, and the cars should also respond faster to the cars in front moving. These quicker reaction times will make traffic lights more efficient with more vehicles being able to pass through before the lights change again. The benefits don’t stop there, junctions and roundabouts should also become more efficient.
Could driving become more accessible?
I believe that autonomous vehicles will do wonders for people with disabilities. Self-driving cars could enable disabled people to gain more freedom, they would no longer need to rely on others for transport.
How would driving licenses work?
We couldn’t remove licenses entirely as there are still things that as the driver you need to know, for example how to check the car to make sure the vehicle is road safe. I believe that a new tier should be added. In the UK we currently have different licenses for manual and automatic cars, a new self-driving only tier could be implemented.
This is a slight side note, but while we are on the topic of licenses, I also believe that regular ones should be re-tested every few years. The reason I think they should be re-tested is it would improve everyone’s safety on the road. It is no secret that as we get older, our vision and our ability to concentrate, react quickly, and judge hazards deteriorate. Our senses worsening is fine in day to day life, but when there a split second judgment required to determine the outcome of life and death situations, this is a significant problem. Re-testing every few years would ensure that everyone with a license is still safe and also keep everyone updated with the latest rules, and knowledge required to drive.
Another way to look at it is like this, drivers of trucks, forklifts, elevated work platforms, planes, trains, and construction machinery all have to either re-take their tests every year or two or are required to participate in yearly training. Why are cars different? Especially when cars account for the majority of injuries and fatalities out of all of them.
Will autonomous vehicles help the environment?
In short, this highly depends on the development of electric vehicles. I am a big advocate of electric vehicles, and I believe that they are the future, but I also know that they have a long way to go before they become mainstream. It is likely that self-driving vehicles will help the environment, or at least not be as bad as petrol and diesel-powered vehicles, this is because it is likely that the vast majority of autonomous vehicles will be electric. Any benefits to the environment won’t be seen for a very long time, as for things to really change, there would need to be widespread adoption of electric vehicles.
In my opinion, the most significant benefits will come from automated electric public transport. Having good, automated, renewable energy powered public transport systems would, in theory, reduce the number of cars on the road, thus reducing the number of vehicles being manufactured, which in turn would lead to fewer pollutants being released into our atmosphere.
Will people lose their jobs?
Yes is the simple answer, but this is very industry dependent. Automisation will allow companies to use select vehicles without any humans in them which therefor spells the end to humans being required in specific areas of certain industries, such as taxis. I have read a lot of concerns about the issue of mass unemployment, and like I say some types of job will be affected more than others.
One of the industries I have seen the most concern for is the trucking industry, as removing humans from the equation would allow everything to become more efficient as the costs would be lower and there would be no need to worry about workers hours. In the UK there are just over 300,000 HGV drivers, in the US there are roughly 3.5 million truck drivers (the largest amount of any country), and around the world, there is estimated to be between 5 and 6 million commercial truck drivers. In total, this is a scary number of people that may lose their jobs. However, I feel the trucking industry is one of the least threatened by autonomous vehicles because we will still need skilled people who know how to hitch trailers properly, perform safety checks, drive moffets, and diagnose issues.
Even with potentially substantial job losses, new jobs will also be created, but I don’t think enough new jobs will be created to offset the effects of all of the ones that will cease to exist.
What about the weather?
Our planets natural elements are one of the self-driving vehicles biggest downfalls, the biggest of which being snow. Anyone who has been somewhere where it snows will know that when it does the lanes on roads quickly become one, and in some places, the road itself can even become indistinguishable amongst the rest of the landscape. Autonomous vehicles rely heavily on lane markings, so how do they work when there aren’t any? Currently, as far as I understand, there doesn’t seem to be a solution to this problem yet. I am not even sure if this is a problem that can be solved, what else is there to tell a vehicle where it can drive other than lane markings?
Conclusion
I have only just touched the surface on the topic of the autonomous vehicle. Self-driving cars are still in their early days, and it will be at least a decade before we really start to see the impact that they will have. That doesn’t mean that we should wait until then to start trying to figure out how to deal with the problems that they will cause, and help to lead companies in the right direction with developing them. It isn’t all problems though, there are also positives, big changes are coming, and it is only natural to be worried, but change isn’t always a bad thing.
