Robot that travels specific distances at measurable speeds

[SciOlyStudent]

I'm currently designing a robot that can travel a certain distance (which I can change quickly and accurately) at a certain speed (so I can predict the time for it to travel that distance).  I was thinking I could use the 50$ robot setup, except add something to measure distance traveled instead of light.  What modifications would I have to make in order to do this? Thanks for any help!:D

[Admin]

As always, I have a few questions to ask:

What is the robot terrain like?
Your budget?
What is your acceptable error rate (absolute minimum distance off)? All sensors have error . . . high acceptable error rates are cheaper and easier

[SciOlyStudent]

What is the robot terrain like?
Your budget?
What is your acceptable error rate (absolute minimum distance off)?

1. The terrain is hard, level and even.
2. Very limited.  This is why i asked if I could just modify the 50$ robot to my needs.
3. Hmmm... well, I'd have to say that I would like a fairly small error rate, maybe a few centimeters or less.
      This is a competition in which my robot has to travel a specified distance which will be announced just prior to the competition (literally minutes), so it has to be a quick adjustment type situation.  Thank you  .

[Admin]

You will need encoders for this task.

You can either make your own, or get this:
http://www.acroname.com/robotics/parts/R238-WW01-KIT.html

[bens]

Does your robot just have to go in a straight line?  It sounds like going at a constant speed doesn't have to be a concern.  Basically all you need is a simple encoder to keep track of your wheels' rotations.  Typically you want to use two sensors per wheel for what's known as "quadrature", which lets you determine the direction the wheel is rotating along with its rotational speed.  If your wheels are always only moving forward, you would just need a single sensor for each wheel.

Actually, if you really just have to go in a straight line on a flat course, you could mechanically link both wheels via a single axle and simply monitor the rotation of one of those wheels with a single sensor and an encoder wheel.

[SciOlyStudent]

Does your robot just have to go in a straight line?  It sounds like going at a constant speed doesn't have to be a concern.

  What you say is true, but in the first post I mentioned the fact that I need to be able to estimate the total time it will take my robot to travel the specified distance.  Hooking the two wheels to one axle and tracking the rotations that way is a good idea.  Just a question about servos, if I can make the robot stop after a certain distance, will the servos automatically stop the robot at that distance?

[Admin]

Just a question about servos, if I can make the robot stop after a certain distance, will the servos automatically stop the robot at that distance?

Depends on many factors . . . what you will end up doing is designing a PID control algorithm, where you test to find the coefficients needed to make your robot stop as fast as you want.

check this out:
http://www.societyofrobots.com/programming_PID.shtml

What you want is a very short settling time (not measured, just minimized, right?). Basically put in the Actuator_Output equation as in the tutorial into your robot, and change the tweaks until your robot does what you like in testing.

If your robot is light weight, you can ignore the acceleration ('I' term).

[SciOlyStudent]

So can I basically use the 50$ robot design except connect the encoders instead of the photoresistors?  Or is there a different setup (circuitry) needed in order for it to accept the encoders?  If possible, is there a cheaper way to make this robot (less than 50$)?

[Admin]

So can I basically use the 50$ robot design except connect the encoders instead of the photoresistors?

yeap . . . but you will need to change the code a bit . . .

as for getting cheaper, you can always make your own encoders . . .