Engineering Stack Exchange is a question and answer site for professionals and students of engineering. It only takes a minute to sign up.
Sign up to join this community
I am confused about the difference between control systems and artificial intelligence, since both techniques can be used for automation.
Please kindly try to give answer in a nutshell, an in depth or detailed answer may cause confusion because I am beginner.
I have already put this question on another SE site at following link but the answer was detailed/in-depth.
https://robotics.stackexchange.com/questions/19319/control-vs-artificial-intelligence
The control system is a device which gathers the instructions (desired behaviour) from some source (e.g G-code, sensors, etc.) and produces signals to the servos (through drivers or a power layer). It is usually quite dumb (compared with AI), and low-latency hard-real-time, however the term does not exclude the usage of AI.
AI implies some sort of advanced computational behaviour:
Those systems are able (for example) to recognize some object/aspect on a video and trigger the consequent actions.
It requires highly intensive computation, with many-cores CPU/GPUs which make it mostly incompatible with low-latency MCU used in control systems.
Note: AI is a quite abused terminology: its specific meaning has evolved significantly and can be applied to a wide range of technologies.
AI is usually considered an input to the control system rather than part of it.
To summarize, a control system could include AI, but AI is not a control system. AI is usually considered an input to the control system rather than part of it.
Robotics is not my specialist field but I will try my best to explain this.
How you control any machine is via the control system. Obviously in Robotics this can be relatively simple or very complex depending on what the robot is doing. Control systems are today usually "controlled" by a program loaded into some sort of electronic interface device like a P.L.C. (Programmable Logic Controller). The P.L.C. will go through a series of steps looking at (information in) and activate outputs as the program requires.
Example if input a + b is present activate output 1
The program is fixed unless you alter the program the device will always perform the same functions according to defined inputs being correct and will only perform duties within the constraints of the program.
AI is similar in that it controls something but it has intelligence, it can learn.
Example if input a + b is present activate output 1 but last time it had no result because the machine had been switched to maintenance mode. So the A.I. self adjusts the program such that if input a + b is present and maintenance mode is not selected activate output 1*
The only problem with A.I is that the intelligence is only as good as the part of the program made to represent that intelligence.
