Image 1) Oldsmobile 442 (Illustration image)
When researchers are creating algorithms and self-operating robots, they need an environment there are strict and certain rules. The traffic environment offers that kind of operational environment. The first AI algorithms used the chess game as the test platform.
Chess offered a good, well-controlled but limited environment for the AI creators. But then the power of computers and things like decentralized computing along with AI-based solutions makes it possible to use a larger environment when ambition grows. The self-driving cars are engineers' dreams and that kind of system requires lots of complicated code.
Making self-flying aircraft is easier than making a self-driving car. The flight altitude of the aircraft must adjust to higher than the highest mountain. And the aircraft will not impact that mountain. The self-driving car must avoid many things like surprisingly starting danger.
Image 2) Delivery robots can use to collect information that researchers need for making self-driving cars.
The self-driving cars require lots of code and many variables. That means the programmer must be accurate. And the code must be error-free. If there are problems with the program code. That means the car can cause a dangerous situation.
So the things like delivery robots used in markets play a big role in this kind of development. Those robots can collect information on what kind of changes and bugs the control code of the systems has. The car pays over 1000 kg, and if it has problems with control algorithms that thing can cause a dangerous situation.
But when we are talking about traffic as the platform for testing and operating the AI, we must understand one thing. The data mass and accuracy of data are the things that are making those systems suitable. If the AI can get information about the beginning and destiny of each vehicle.
And information about predicted breaks, etc. that allows the system to operate as an entirety. The system can make holes in car rows in the motorways and adjust the driving speed. But there are always problems with privacy.
Information plays a big role in the AI-controlled entireties. Drone swarms require information about the positions of their members. The algorithms that are controlling traffic are basing a similar hierarchy to the drone swarms.
The hierarchy in the control protocol is simple.
1) The system that operates the entirety interacts with
2) Individual systems. Those individual systems can be the systems that are operating cars. But they can be infrastructural systems.
When some drawbridge rises the system stops traffic. And the operating system can also interact with the ships and fit its speed so that passing the bridge is comfortable. This kind of entirety where multi-level systems interact makes the traffic smooth and flexible. But there is lots of work with those systems.
Comments
Post a Comment