We were given this as a class project, with only 4 weeks to produce a working prototype. While a lot of people had some really creative ideas, I knew that only the simplest, most straightforward design had any chance of being successful.
So my first suggestion is think simple.
Now robots are called as such because they sense, actuate, blah blah blah. Forget all that for now. Go to a 3 year old and ask him what kind of vehicle can climb stairs. He will say 'a tank'.
Yep, we built a small tank. The length was such that it easily spanned two steps. The 'treads' were bicycle chains riding on sprockets. Motors (with smaller sprocket gears attached) drove the chain forward or backward. We installed roller blade bearings between the idling sprockets and the fixed axles. To some of the chain links, we attached metal barstock (for grip).
The thing was like two chainsaws. It didn't need a microprocessor; you just had to turn it on and get the hell out of the way.
Now once you get something that can work without 'intelligence' so to speak, you can then think about adding a microprocessor and making it a true robot. There are many possibilities here; use your imagination.
2006-12-21 05:14:22
·
answer #1
·
answered by Anonymous
·
0⤊
0⤋
Design an appropriate frame that can span stairs. The frame will need some kind of legs that can span the stairs.And an appropriate power supply and electrical package.
Design servo controls that can position and move the motors you will put at each joint in the legs. Be sure to do a full stability analysis of all your feedback loops.
Design an optical recognition system that images the stairs.
Design interfaces between the servo loops, the optical system and your controller.
Write a program that takes the input form the optical system and determines how to move the position of the servos to climb the stairs.
Should only take a few years if you're good.
2006-12-21 12:26:20
·
answer #2
·
answered by Gene 7
·
0⤊
0⤋