Ideas Needed and Directions too!

[mazlano27]

Hi, im looking to make a robot, which has to follow a pre programmed path. We cant alter the track in so no tapes are allowed to be pasted. I need to make a robot which can get from point a to point b and transfer a load! Where should i start?

[Resilient]

It sounds like you should start at point a.

Seriously though, what research have you done thus far? If you haven't done any, then that is where you should start. Reading about robots that do similar things to what you want to do.

[newInRobotics]

Are You saying that path has to be followed without any references to the outside world? If that's the case, then read about dead reckoning.

[mazlano27]

i have done plenty of research, and im looking around for things like preprogrammed motor driven robots which i can construct on my one. And i have been given the blue prints for the track and have to do guide the device through without touching it or without any RC. This is for my engineering course, the constriants are also pretty strict. i have an idea of what i have to do, just dont know how. for example how am i to pre programme a motor, which programming language is utilised etc...

[newInRobotics]

And i have been given the blue prints for the track and have to do guide the device through without touching it or without any RC.

Does the track have walls? If so, proximity detection is the way forward.

for example how am i to pre programme a motor, which programming language is utilised etc...

You don't pre-program motor, You program microcontroller to drive motor in a way You want it to be driven. Language choice depends on the microcontroller You want to use.

[Soeren]

Hi,

i have done plenty of research, [...]

Judging from what else you wrote (and the simple fact that you ask), I don't think so, but perhaps "plenty" means something else to you, than it does to me

I think it wouldn't hurt with a wee bit more research.
Building something simple would be a good way to start - you'll overcome some obstacles much easier on something simple and this will help you when working on the real deal.

[Resilient]

What level of engineering student are you? And again, what research have you done? Show us where you are confused in the things you are reading and maybe we can help.

But getting a sense of the level of course will help me figure out what caliber of robot they are looking for.

You could get something very very basic going by just building the $50 robot and using timers to control the motors. It is almost guaranteed to fail but it sounds like you don't have much experience so this may be one of those exercises where the instructor is trying to make a point about how awful dead reckoning is.

But if there are more specifications you are not telling us, like markings of the course, then you can do more.

But again, show us the things you have looked at in your research of this project, tell us what you don't understand and we can probably do more for you.

[jkerns]

There are a couple options for navigation - but it sounds like you need to program a robot to operate "open loop" - it tries to measure how far and in what directions it has moved so it know which way to go from where it thinks it is at.

So you need to know two things. What direction it is pointing, and how far you have gone.

How far would come from keeping track of how many times each wheel has turned - right? So you need encoders of some type on the wheels. What direction implies either a compass or starting off in a known direction and keeping track of the difference in wheel speeds.

It's just math beyond that.

Does that get you started?