Step climbing mini robot

We are a small group of engineering (electronics, computer science and mechanical eng) students who are trying to design a small robot (around the
size of a 3 years old kid) that can navigate both on smooth indoor environment, on walking paths of a uni. campus and also climb steps. We want this robot as a personal companion which can go around with its owner. We are trying to build this not only for fun, but we beleive all of us will learn alot from the practical side of the project.
So we are asking all, if you can contribute with your ideas toward solution of a particular issue that you think it is important to solve for the success of the project. The idea of asking this question is to create some sort of brain storming session so that everyone who is interested in step climbing mini robots can learn from others.
Can some one suggest which sort of motion mechanism we should start with?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Start by looking around for other robots that climb steps and travel on walking paths. How many do you see ? - not many. This isn't an easy problem to solve, particularly for a smallish robot where the obstacles will seem large. In general the human environment is designed for walkers, with token offerings for wheelchair users. A quick search for walking robots will show you that they are not easy either ....
So, a few hints, without any guarantees.
* 'Small' is not necessarily good news when the obstacles are relatively big.
* Walking is hard work and takes a lot of power and control. It also tends to be very slow - perhaps not the ideal companion ?
I'd start with something that could cope with rougher ground - perhaps taking inspiration from wheelchair design
Best of luck !
Dave
<Stair> wrote in message> We are a small group of engineering (electronics, computer science and

the
want
solution
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Asimo can climb stairs, but then Honda put maybe $100,000,000 or more into its development, plus it needs an external (super)computer for control.
http://world.honda.com/ASIMO /
On a smaller [non-android] note, Rhex is one of the few "walkers" than can go easily over broken terrain, climb obstacles, etc.
http://www.rhex.net /
- dan michaels www.oricomtech.com ===========================
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Stair wrote:

I wish you luck. This sounds like a very complex problem.
First, allowing a robot to go outside can be somewhat difficult and potentially dangerous if it is powerful. What you are designing is something like what I'm thinking about for my next project.
I'm basing this on an electric wheelchair. This should give me a robot about the size of a medium dog. This will probably weigh about 150-200 lb and have enough tourqe to move up a set of stairs.
This robot will be carefully monitored in the beginning. I figure it has enough power to seriously damage my house, not to mention my dog, myself, my wife, etc... Visitors will only get a *very* short and specific demo.
As to stairs, I'm thinking of a basic differential drive robot with the wheels in the middle. There would be casters on either end for normal use.
In the front there would be an extra set of powered whegs or treads (I haven't experimented to see which would be better). Their main job is to help push the front of the robot up the stairs, with the wheels providing the main power. Eventually the whegs/treads would be in a recessed compartment until needed, or perhaps used as a bumper.
These are just thoughts, and nothing is even in design-layout yet. -- D. Jay Newman http://enerd.ws/robots /
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.