It seems somewhat of a nerd’s dream, but how likely is it to see a robot driving a car? We have the idea of small robotic toy cars and larger self-driving cars, so you may have thought that it was about twenty or so years in the future? But let’s take it slowly.
Why are robots seen as so different from an AI self-driving car?
There’s definitely something about them not looking human despite the fact that robots can look like anything.
There are a few self-defeating aspects and somewhat silly parts about the principles here. How would you even begin to get a robot to open the door or wind down the windows? But assume that you could get a walking robot in the driving seat and driving, what happens next?
You have a unit you can transfer from car to car for one. You can also see the driver – though you could say the same with a dummy, or a hologram, or even someone on the front seat who isn’t driving? Instead of improving the car you “only” need to improve the robot, rather than the self-drive of any specific car.
It’s also something which is using the steering wheel, so you still need the “old” or conventional parts of the car, not just the steering wheel but also the clutch, the pedals, the horn, whatever.
Another helpful factor is that the robot is able to leave the car, which it is possible to utilize by getting the robot to drive in enemy territory, make the robot leave the car and then blow the car up.
Now the bad points.
Just as with a car a robot can fail, it would need replacement bits. And how many robot technicians live near you?
Could a robot even sit in a human driver’s seat? More likely you’re looking into a whole new, more expensive interior.
No one really wants to see a robot driving a car, we only think we do.
There’s also the possibility of hacking if you’re talking about driving a car from afar. It is after all a computer.
How easy is it to teach a robot to drive? Can it respond to stimuli? Can it avoid obstacles? There’s even the weight problem when the robot attempts to move from car to car.
It’s also a painstaking process; it’s just quicker to build the AI tech into the car itself. But could this be seen as old fashioned in the future as all previous ways of driving are dropped.
So it’s a Terrible Idea? Well…Not exactly
Amazingly though, the start-up company IVObility developed a robot driver which saw the road ahead in 2020. Despite being lightweight enough, resembling a kind of stick figure, it wasn’t a robot as such since it wasn’t autonomous and relied on remote control, so more of a magic trick. It should be able to build a robot which learns from demonstration, but we’re not there yet.
The robot which drives a car may unfortunately lead to life-or-death situations. This is the true worry about the whole process.