How effective will self-driving cars be in the current road system?

This is the future of AVs… or is it?

Machines are becoming more and more prominent in our lives as time moves on. As this happens, many companies are attempting to make fully autonomous cars 🚗. One of the main goals of autonomous vehicles is to optimize road travel, effectively eliminating human error. Many people visualize the future road system to be similar to the gif above, since the cars can communicate with each other at incredulous speeds 💨. However, this would require all cars on the road to be autonomous. So, here’s my question: how well will self-driving cars fit in with our current road system?

The current road system

Pros:

Cons:

The current road system was developed for human drivers so that the task of driving becomes easier and safer. However, if all cars were self-driving, many features would be unnecessary, and might actually impede the functionality of the AVs.

Yet, the potential that autonomous cars have is immense. The prospect of no traffic, ensured safety, and more free time entices everyone. The chore of driving would be completed by a machine, so anyone would be eligible to ride. Another benefit is that emissions would be lower because of ridesharing and optimized driving. If everyone accepted autonomous cars, we wouldn’t have to deal with traffic jams.

This all sounds great! So what’s the problem?

Well, Is it possible?

If you’re an optimist, anything is possible 😄! However, when you consider the realism of it all, the benefits are exaggerated 😬.

First of all, the general public doesn’t trust autonomous entities. More specifically, about 75% of Americans [surveyed] would rather drive than ride in an autonomous vehicle. No matter how many demonstrations and data points there are that support the safety of AVs, there will always be a group that does not trust the technology. The question is, how large will that group be? As of today, the group of skeptics is larger than the companies want them to be. Of course, they have their reasons, such as the 2018 Uber self-driving car accident, which was detrimental to the reputation of AV safety.

Maybe this is why people don’t trust machines

Also, if we wanted to fully optimize road transportation, then the whole road system would need to be reworked. All road signs, barriers, and even lanes today, are not effective for autonomous cars. Even if every car on the road is autonomous, they would still need to follow the road rules we implemented (for human drivers), limiting their efficiency. Changing the system would take too much time and resources, and many people may disagree with it.

Now, consider how many companies are working on autonomous vehicles: Waymo, Uber, Nvidia, a vast number of automobile companies. All cars would serve the same purpose, but operate differently because of the different software. This could create biases: some cars could be programmed to be more aggressive than others, some could be made to drive faster, etc. This could start a battle of brands. Just imagine: “Yeah, the ___’s cars drive well, but ___’s are faster!”. This would cause discrepancy and unfairness on the road, which is a trait displayed by human drivers.

Finally, in my opinion, there are too many factors for the general public to fully trust and accept self-driving cars. There is the risk of malware/hackers manipulating the software, or bugs in the code. There is also the risk of pedestrians acting unpredictably, but there always has been this.

My opinion is that self-driving cars will end up on the road, but I don’t think the goal will be to have all cars on the road be autonomous. The goal should be to prove their effectiveness and increase safety on the road. Then, who knows? If the resources were available, it would be great to have all cars self-driving. You just have to prove to the skeptics first.

Current Solutions:

Just because there are some downsides to the current situation doesn’t mean that we will never see self-driving cars on the road. In fact, companies are working around these problems as you read this.

Security is a crucial feature of a self-driving car. Many have expressed fears of self-driving cars being hacked; there have been situations where control over the car is lost. However, self-driving cars are becoming quite secure, and rather than keeping their information confidential, companies are sharing their findings in this field. Through techniques like localization, removing unnecessary components and not allowing inbounding connections, security is made stronger.

The road and the environment around it can be unpredictable; this has always been the case. When an accident occurs with self-driving cars, it severely affects their reputation. However, many people don’t compare the numbers. A study of vehicle incidents in California found that humans were the main cause of self-driving car accidents.

The study, which spans 2014 through 2018, found that when the self-driving cars were in autonomous mode and driving on their own, 38 incidents occurred while moving. In all but one of those cases, the accidents were caused by humans.

To promote safety, manufacturers are adding extra safety features to autonomous cars (ex. cars give extra space to objects/pedestrians). In fact, the larger problem is that self-driving cars can be too cautious; riders have complained about long waits to turn and slower speeds.

As for the current road infrastructure, autonomous cars have the ability to monitor the environment with great accuracy. Tools such as street sign recognition, lane sensing and LiDAR help self-driving cars adapt and fit into today’s roads. If autonomous cars can prove to be effective, then maybe some rules will be put in place for their advantage (self-driving car lane, maybe?)

Conclusion

Self-driving cars are nowhere near perfect, but they are rapidly evolving. The future is unknown; experts have made many predictions as to when they will be commercially available. When they do become available, the change will be radical. Ridesharing, the ability to do different activities in the vehicle, and more benefits will definitely attract people. However, will they last, or is the excitement superficial? I will, once again, say that only time can tell.

Student | TKS Alumni