Subject:
|
Re: Seeking Strategies on Humanoid Robots/Walkers
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Thu, 21 Jan 1999 17:35:49 GMT
|
Viewed:
|
4786 times
|
| |
|
|
Patrick Gili wrote in message ...
> > I seem to recall that noone has succesfully solved the problem of
> > artificial biped locomotion even in traditional robotics, let alone
>
> This problem has been solved with both fuzzy logic and neural
> network control systems. I believe that a traditional control
> system approach to this problem has yielded poor results.
Actually, they can be built with *relatively* simple, active-feedback
systems for the balance and control.
Simple legged robots (one leg is actually easier than two to control) have
been around since ~1980.
See MIT's "leg" laboratory for pictures and videos!
http://www.ai.mit.edu/projects/leglab/robots/robots.html
> No, I don't think anyone will be solving this problem using
> Mindstorms. There is a whole list of limitations:
>
> - slow processor
> - wrong processor architecture
> - wrong software tools
> - inadequate sensors
> - inadequate actuators
The biggest Lego limitation is in the acutators and sensors. Sensing the
"delicate shifts in balance" _is_ difficult, as well as making delicate
manuvers to correct for those shifts--especially when you scale things down
to managable Lego sizes.
- Izzy
|
|
Message is in Reply To:
 | | Re: Seeking Strategies on Humanoid Robots/Walkers
|
| Jasper Janssen wrote in message <36a3ca48.45364286@l...et.com>... (...) This problem has been solved with both fuzzy logic and neural network control systems. I believe that a traditional control system approach to this problem has yielded poor (...) (28 years ago, 21-Jan-99, to lugnet.robotics)
|
20 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
