Science fiction has a way of turning to reality, and now we are in its midst. Startups like Waymo are operating driverless taxis without anyone behind the wheel. There are some places that it is running, but not all places yet. We know these AI technologies are almost flawless and not subject to the human nut, behind the wheel.
The big question is whether a self-driving robocar has your full confidence, that if it can handle the erratic drivers outside. Can it get anywhere in one piece? Least in mind is the passenger, who is panicking and screaming in the back, with no way to take over. That is, how it works, and not everyone trusts new technology.
One incident that caused a furor, is the accidental death of a pedestrian due to Waymo. All because the robo-car did not know how to react. That is exactly the problem faced by robo-cars that need to be taught how to react. Another thing is that the Lidar and other sensor are limited, not like human senses. Their limitation might place the rider in a rock and a hard place. For the meantime, robo-cars are having improvements for dealing with the situation in human terms. Now, here are the kinds of autonomous driving that we have now before we achieve full self-driving.
1. Driver Assist features
The purpose of driver-assist features is to help manage driving automatically when needed. Most accidents happen because the human driver lost focus. For robo-cars, this is easier because they got more sensors, with automatic systems that are always peeled. The features self-driving cars to avoid collisions are collision and pedestrian warnings, with an autonomous emergency brake that activated on its own. Other features active cruise control keeps the car at a safe distance.
2. Semi-Automatic Driving
For the most part, the driver still overseers this function as the car runs. Unlike driver-assist features, this will control steering already but not take over. Important functions are unguided steering, lane changes, and how to handle congestion to some degree. For Waymo the Lidar mount has more sensors than ordinary cars. But the drive must be alert at all times, no matter what.
Now, cars will drive autonomously and the driver will cede total control to the car. This kind of programming allows driving for long distances, but the driver should not let the AI do everything. This third stage of self-driving has limits. So, be safe and be alert, to take over anytime.
4. Full Automatic Driving or L4
Reaching this level means that the car can drive itself. There is still the driver's seat, and the human driver can take over if needed. The complexity of it is demonstrated in Waymo car, but everything depends on the sensor or the Lidar mount. It can handle most situations to correct itself and even some difficult situations.
5. Full Robo-car Function or L5
At level 5, this is where full automation is reached, and the AI is at its best. The car can handle most driving tasks better than human drivers. There might not be a driver's seat any more at all. This is the goal of autonomous driving. It can handle itself and adapt to any situation thrown at it.
For now, the self-driving robo-cars are yet to be fully autonomous, but we're reaching level 5 fast. With the coming technologies, we may soon be forced to decide if driving our own cars is needed at all.
Related Article: The Path to Autonomous Driving