Swarm Robotics Prototype 1.0: Why prototype 0.0 will fail...

Hi all. I hope you tried out the last project i suggested to get some insight into swarm robotics. Now all those who did mostly know why I am writing this post. The swarm robotics algorithm that we discussed in the last post fails because of a very practical reason. The second robot that was following the first does not have to turn when the first does, it has to turn a 'little later' And this is where we come across a very important aspect of robotics, that of calibration.

Calibration is a process where we relate how much time a parameter needs to be supplied to a component of the robot so that we have control over its action. As an example, we need to know how long a dc motor needs to be kept on so that the robot moves by say one meter. The process of calibration requires experimental tests, as a lot of factors like texture of the surface, weight on the bot etc. come into the picture.

I am writing here an extended version of the swarm robotics problem statement we discussed in the last post that will make the problem statement more accurate. The robot that is following the first robot must EXACTLY follow the line. I would suggest something called a proximity sensor that will make our task easier. However how to use it and a little more detailed discussion on calibration will be done in the next post.