Artificial Intelligence in automobiles: Is it redefining the way we drive, or expose us to potential safety hazard?

Over the past couple of years, the scope of technology in cars has expanded significantly. It was not too long ago when Tesla flirted with semi-autonomous driving, and other manufacturers tested their autonomous technology. Now, not many manufacturers offer a full-blown autopilot as Tesla does,  but if one is aware of the latest industry trends and technology that’s being utilised for the mass-produced cars, you’ll know the latest tech, such as ADAS, which is built on artificial intelligence technology, is pretty close to a full-fledged autopilot experience.

Automotive Technology such as ADAS, which includes AI driver assists such as automatic braking systems, lane keeping assist, lane departure assist, driver monitoring systems, pedestrian detection, and protection, the list just goes on and on, and not to mention your traditional driver aids such as traction control and vehicle stability program, which surely made automobiles a lot safer to drive.

But my qualm with AI integration in automobile technology, which essentially enables the car to drive itself, is, does it make driving safer? Or minimising the human element has actually made driving more complicated and exposed us to potentially new safety hazards arising from excessive dependency on driving aids. It might seem counterproductive that features that are meant to make driving safe will also be the reason to make things unsafe, but there are a few instances that I have come up with that might pose a safety risk while driving, dependent on these driver aids.

Over-reliance on Forward collision avoidance systems

It’s been a while since some high-end cars got these systems, which essentially detect the distance between the vehicle ahead and the speed of the host vehicle and determine whether a collision is about to take place between the two vehicles, and apply necessary measures to avoid the predicted circumstances or at the very least minimise the damage in the event of the predicted incident. It does this by either alerting the driver through some in-cabin notification system or, in some highly sophisticated systems, it detects and analyses the situation and applies the brake automatically without any input from the driver.

So, my concern is what if the system fails when it’s actually needed, and the driver has gotten over-reliant on the system and isn’t paying the same attention to the road he used to? I mean, at the end of the day, the features work on a network of connected sensors and processors that are supposed to measure and calculate millions of datasets to their final approximation, and the sensors or the processors are electrical units that are prone to failure. Or what if the system fails to detect the car that’s behind, and inadvertently brake checks the car behind, causing one incident to avoid another.

Autonomous driving 

Recently, I have been in a car that was equipped with level 2 autonomous driving,  and got the chance to spend a decent amount of time behind the wheel, and explore the full potential of the autonomous features that were on offer.  With all my apprehensions about the autonomous systems, I must admit that it did make a considerable difference in my driving experience, sure certains beeps and bongs were annoying at first but as my journey shifted towards the stretcedh highways with miles to cover,  the autonomous features came in very handy, i was much more relaxed and as hours went on i experienced less fatigue than usual.

As relaxed as I was with the car doing most of the work for me, a daunting thought crept into my mind. I wasn’t as vigilant as I would be if it were any other car, and as good a job as the system was doing up until then, it made me paranoid. What if the system fails when I need it the most? Or what if I get too heedless while driving, or I rely too much on the system, and when a situation arises, it might be too much for the system to compute and avoid the situation, which could have been easily avoided if a human were paying that extra bit of attention on the road. 

Sure, all this is conjecture on my part, and no manufacturer that offers autonomous driving ever encourages its users to completely let go and relax while the car is driving itself. What do you think, will the advancement in Artificial Intelligence and machine learning algorithms enable us ever to see an autonomous system that will completely replace human input and be just as safe?

0