 | | tracking robot position
|
|
Hi all! I had an idea last night, and I think something good might come out of it, with so many smart people around! *s* I need a way to track the position of my robot, and I realized that attaching the PC mouse to it might do the trick! Without (...) (25 years ago, 11-Nov-99, to lugnet.robotics)
|
| |
| | Re: tracking robot position
|
|
Using two rotation sensors should do the job nicely, at a higher cost however. But you can connect them directly to the RCX. About that : did anyone build a reliable (and cheaper) rotation sensor replacement? Why not using a lego wheel with holes (...) (25 years ago, 11-Nov-99, to lugnet.robotics)
|
| |
| | RE: tracking robot position
|
|
This is an interesting idea, but I think you will have a problem common to most dead-reckoning navigation solutions and that is that wheel slip, mouse ball slip, and the differential slip between the two (i.e., if the robot wheel slips, the mouse (...) (25 years ago, 11-Nov-99, to lugnet.robotics)
|
| |
| | Re: tracking robot position
|
|
(...) Of course! But for some applicvations, that isn't a big problem - and if it works it is cheaper and simpler than other methods at the same precision (I think) As a matter of fact, I am working on a way to use a pattern printed on the floor and (...) (25 years ago, 11-Nov-99, to lugnet.robotics)
|
| |
| | Re: tracking robot position
|
|
(...) with (...) attaching (...) PC (...) Hi Vlad, the main problem I see in using a mouse for dead reckoning is that you cannot register the rotation on its own vertical axis. So this solution is, by principle, not working. If you're not convinced, (...) (25 years ago, 11-Nov-99, to lugnet.robotics)
|
| |
| | Re: tracking robot position
|
|
Vlad Dumitrescu wrote in message ... (...) with (...) attaching (...) PC (...) Use a cordless mouse. They use radio and have a range of several metres. (...) It's time Lego came up with a GPS sensor for the RCX ;-) (...) extension (...) such (...) (25 years ago, 11-Nov-99, to lugnet.robotics)
|
| |
| | Re: tracking robot position
|
|
(...) Hi! If the mouse ball is placed at a distance from the rotation center of the robot, then it will never get into a situation where it will rotate around its own axis. Of course, the software has to perform some calculations, but it's always a (...) (25 years ago, 12-Nov-99, to lugnet.robotics)
|
| |
| | Re: tracking robot position
|
|
(...) My initial thought was to use an old discarded mouse, at cost 0... ;-) (...) There are some of those, using optic fibre from the Extreme Creatures kit, for example. /Vlad (25 years ago, 12-Nov-99, to lugnet.robotics)
|
| |
| | Re: tracking robot position
|
|
This isn't suitable for Vlad's original problem, but I do want to point out that motors can be adapted as cheap rotation sensors (actually rotational velocity sensors, which is the first deriviative of what a normal rotation sensor gives you) by (...) (25 years ago, 13-Nov-99, to lugnet.robotics)
|
