A robot that is programmed to react to external stimuli

[Tabaristiio]

A robot that is programmed to react to external stimuli, what factors determine the speed at which it detects / recognizes the stimuli?

Suppose this robot is programmed to respond to various different external stimuli, such as clicking a button when a color changes, clicking a button when something appears in a blank screen, evading a punch / kick etc.

This question focuses purely on the stimuli recognition / detection speed and not the actual physical action / reaction speed.

Say I wanted to program this robot to detect / recognize various different stimuli at 0.1 milliseconds, what has to be done to achieve this?

When I say various different external stimuli, I mean it being multi - situational. So this robot should detect / recognize various different external stimuli in 0.1 millisecond.

For example, detecting a punch / kick from 5 inches away at 0.1 millisecond. Detecting a color change in a computer screen in 0.1 millisecond. Detecting a pistol making sound (like in 100 sprints) in 0.1 millisecond. Detecting something appearing in a computer screen in 0.1 millisecond etc.

I've learned that using photo diodes can be useful for detecting color changes very fast. However, can it be equally helpful for detecting other external stimuli such as detecting a random piston sound, punches, kicks, etc. or is it purely for detecting color changes?

Is there something that can be used to improve detection speed for all external stimuli or is it specific to different ones?

Please note that my question is purely focusing on visual and auditory external stimuli and detection speed.

Another important note is that my question is purely driven out of curiosity. It's not a professional job or anything alike. I'm just trying to learn this out of curiosity.

Thanks in advanced!

 

[Tabaristiio]

I just wanted to add one more thing. Even if the specifications I put forward isn't possible, such as a reaction time of 0.1 second, then please state what has to be done to the robot for it to get the fastest reaction time that is possible. Whether it's 0.10 milliseconds, or 0.50 etc.

 

[willem2]

You're limited by how fast the information is sent to the screen, and is received from the camera.
A LCD monitor doesn't flicker, but the screen is still only refreshed 60 times/second, so that takes >16ms.
A camera won't send one image faster, unless you have an extremely expensive one. Then you'll need software to recognize what's happening in the image.

 

[JBA]

Ignoring the limitations of the signal source, for an image the time required is based upon the speed of light (generally a sufficient speed in other than astronomical distances), the internal signal transfer time and the processing time of the robot detection system; but, for audible inputs the speed of sound in the atmosphere related to the distance between the source signal; and the robot detector could be factor as well.
In the case of multiple sources of inputsdevoted sensors and multiple parallel processors could be required as well since most data acquisition systems utilize multiplexing which, similar to the refreshing time mentioned above, is a sequential sensor data input reading process.