With the constant advancement of artificial intelligence in our modern world, it’s no surprise that it’s popping up in numerous industries, including the automotive industry. While AI can be used in car design and production, it is now present in vehicle functions as well.
But is putting artificial intelligence in cars really a good idea, or is it just a disaster just waiting to happen?
How is AI used in cars?
While there aren’t many automated, self-driving cars on the roads yet, artificial intelligence is still present in certain parts of cars. Currently, artificial intelligence is most commonly used in electric vehicles (EVs). This includes not only electric cars, but also electric motorcycles and electric vans. Some electric vehicle manufacturers, such as Tesla, have a clear focus on AI integration.
What specific role does AI play here and what features could be added to modern cars?
1. Accident prevention
According to the Centers for Disease Control and Prevention, 1.35 million people die in car accidents worldwide every year. A shocking 3,700 people lose their lives every day. Road accidents are all too common, but artificial intelligence may be able to help prevent them.
Accident prevention is a key element in many of today’s most advanced cars. For example, Tesla has a feature called Collision Avoidance Assist that offers collision warnings, emergency braking, and obstacle-aware acceleration. In addition, autonomous emergency braking has been used in commercial vehicles and passenger cars since 2013 and will be mandatory for all cars sold in Europe after July 2022.
As part of its collision avoidance capabilities, AI may be engineered to one day monitor and analyze a driver’s behavior on the road to determine common mistakes they make while driving. A driver may regularly brake suddenly or not use the turn signal when turning. These habits can be captured by the AI and used to better prevent the driver from having an accident.
2. Steering and braking assistance
Roads are not always easy to navigate, especially when the weather conditions are bad. However, with the help of AI, driving in more treacherous scenarios could be simplified. AI can come in handy when you need to steer through thick fog, heavy rain or rough terrain.
Additionally, the AI’s ability to recognize hazards and make real-time adjustments can help with braking assistance. You may be approaching a vehicle or traffic light a little too quickly, or the current conditions may require a significantly longer braking distance. With AI, your vehicle could be braked more gently and safely to avoid accidents.
Tesla is pioneering the use of AI to make cars drive autonomously. In 2014 Autopilot was introduced, a semi-autonomous driving system that has some AI-like properties. Since 2020, a more advanced (and even more AI-dependent) version called “Full Self-Driving” has been tested and is currently in public beta. Even if it has led to different opinions due to its very different performance, the fact that it is based on a neural network that actually allows the system to “learn” is a remarkable achievement.
3. Automated Parking
Parking can be a real challenge. Many drivers dent or scratch their vehicle when attempting to maneuver it into a tight space. Therefore, an AI-assisted feature that assists you in parking can be a huge benefit — making parking safer not only for the driver, but also for pedestrians and other people in parked cars.
While many vehicles already have parking sensors that tell you the distance to an obstacle, some also use their exterior cameras in conjunction with their sensor array to actually park, and the driver doesn’t have to touch any of the controls.
In the case of Tesla, this function is called “Autopark” and automatically places the vehicle in a parallel or perpendicular parking space. However, when self-parking a Tesla with Autopark, AI does not play a role. Company CEO Elon Musk announced in 2022 that Tesla’s AI teams are working to improve this feature, so we may see further improvements once this feature becomes AI-powered.
Possible future uses of AI in cars
While the current development of AI applications in cars is already impressive, it is unlikely to stay that way. In the future, AI may be used in cars for many more purposes, including:
Drive automatically instead of a human. Intelligent communication with other vehicles. Conducting driving tests. Determining whether older people are fit to drive.
All of these features have the potential to improve driving safety and comfort. However, before we dive headfirst into the world of AI-assisted driving, there are a few important factors to consider.
Can the AI in our vehicles be trusted?
While the above uses of AI could do a lot of good, there are also a number of risks that should be considered with AI in automobiles, starting with technical failures.
Because AI is a form of technology, it is prone to system failures. While this can happen in a car without AI, relying on intelligent machines to protect us when the technology itself fails could prove dangerous. Suppose your car’s AI-powered hazard detection feature isn’t working properly while you’re driving. If you rely on your car to warn you of hazards, you may need to be less careful to look for hazards yourself. So if there is a system error that you are not aware of, you can easily put yourself in danger on the road.
In the case of self-driving cars, the risks may already be obvious to some. The risk of a self-driving car malfunctioning and causing accidents instead of preventing them is very real, especially while self-driving systems are still in the experimental stages (like Tesla’s Full Self-Driving Beta feature). The National Law Review found that there are 9.1 accidents with self-driving cars per million miles driven, which is more than double the rate for regular cars. There are many car models that offer self-driving capabilities, so the risks need to be considered.
Self-driving cars can also bring other potential problems, such as faulty navigation. If you set a destination and the navigation system fails, the self-driving feature may take you to a different location, unless the system has built-in failover to prevent you from being driven off a cliff or into a lake.
A Stanford Engineering article examined the proficiency and safety of AI, with the author stating that:
[Self-driving cars] They have a harder time recognizing, classifying, and responding to environmental information than their human counterparts. This includes how to accurately recognize and respond to unexpected challenges such as a large obstacle in the road or a child scurrying between parked cars.
This raises a crucial point about how far one can trust the AI with full autonomy. The same article also wrote that “designers, engineers and regulators must assume that autonomous technology is not, and may never be, perfect, and that some interaction between the system and the driver will always be required.” If such considerations are not made can cause many problems and put human users at risk.
Tesla’s Elon Musk was very optimistic when his company launched “Full Self-Driving,” predicting that within a few years the manufacturer’s vehicles would be driving autonomously, without the need for human intervention or even the need for physical controls. However, he and his team have found that the task of making cars self-driving, even with AI and some of the world’s most powerful computers, is proving to be a far more difficult and time-consuming task than first thought.
Increasing reliance on AI for driving can also lead to an overall decline in driver skill and responsibility. We all have to pass a driving test before we can hit the road independently. However, these checks and driver requirements overall could relax as automated driving becomes commonplace. After all, if a car can drive for you without any problems, then what’s the point of learning all the traffic rules?
AI-powered cars have their pros and cons
We have already seen how AI can be useful in many ways both inside and outside the automotive industry. But it’s important to recognize that there are many ways that things can go wrong when we rely on this technology to travel, and improvements will need to be made before we can fully trust it. Who knows, one day we might all be carefree in our own autonomous cars.