Newsgroups: comp.ai.genetic,comp.ai.alife,comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!das-news.harvard.edu!news2.near.net!MathWorks.Com!yeshua.marcam.com!charnel.ecst.csuchico.edu!csusac!csus.edu!netcom.com!park
From: park@netcom.com (Bill Park)
Subject: Re: Make a BattleMech Stand Up!
Message-ID: <parkCxH8pF.J7D@netcom.com>
Followup-To: comp.ai.neural-nets,comp.ai.alife
Cc: milo@MCS.COM
Organization: Netcom Online Communications Services (408-241-9760 login: guest)
References: <377rau$bu8@Venus.mcs.com>
Date: Mon, 10 Oct 1994 21:51:14 GMT
Lines: 51
Xref: glinda.oz.cs.cmu.edu comp.ai.genetic:3978 comp.ai.alife:1114 comp.ai.neural-nets:19373

In article <377rau$bu8@Venus.mcs.com> milo@MCS.COM (Greg Corson) writes:

> We're currently working on the next generation hardware/software
> system now and I have a small problem.  When one of these BattleMechs
> gets shot, or trips over something, it should fall down (obviously).
> This is easy to handle with physics based modeling so it looks good.
> The trick is, once you fall down it would be nice if you could get
> back up again in a way that looks believable regardless of what
> position you landed in.
> 
> ... What I'm wondering is if anyone has any ideas on how you might use
> alife, genetic or neural nets to do this.  
> 
> ... Assuming this works out, the next step would be to come up with a 
> system that evolves a walking/running gait and makes the mech
> recover it's balance when pushed by a weapon's hit.  
> 
> Greg Corson
> Chief Software Engineer
> Virtual World Entertainments, Inc.
> (312) 243-6515
> milo@mcs.com

For my Ph. D. thesis in control systems engineering some years ago, I
simulated a four-legged walking robot and its on-board control system.
I developed algorithms for the controller that successfully allowed it
to keep its balance, walk over irregular terrain blind, maintain its
balance on a slippery hill, hobble along with one leg broken, and
automatically change gates as it walked at different speeds.  It would
trip and fall if you gave it too difficult a problem to solve.

The vehicle was operated by a "driver" who gave it a speed and a yaw
rate.  The robot figured out what to do with its legs completely
automatically.

Although I never implemented an algorithm to get back up after a fall,
the general principles used in the controller should be able to handle
that problem, too, unless it "turned turtle" (no arms or legs or tails
able to reach the ground).  My algorithms could also handle any number
of legs, if you want exotic-looking machines.

If you would like to see the simulation in action, I have a Super-8 mm
film that I shot directly off the display screen.

Right now, my main line of business is something entirely different --
manufacturing applications of neural networks -- but It would be fun
to do some contract consulting work in computer-based entertainment.

Bill Park
=========
-- 
Grandpaw Bill's High Technology Consulting & Live Bait, Inc.
