Newsgroups: comp.robotics
Path: brunix!uunet!usc!elroy.jpl.nasa.gov!news
From: gat@robotics.jpl.nasa.gov (Erann Gat)
Subject: Re: follow-up to posting (non-holonomic motion planners)
Message-ID: <1992Mar31.220240.18912@elroy.jpl.nasa.gov>
Sender: news@elroy.jpl.nasa.gov (Usenet)
Nntp-Posting-Host: robotics.jpl.nasa.gov
Organization: Jet Propulsion Laboratory
References: <1992Mar24.191832.5682@clouso.crim.ca>
Date: Tue, 31 Mar 1992 22:02:40 GMT
Lines: 32

In article <1992Mar24.191832.5682@clouso.crim.ca> freedman@crim.ca (Paul Freedman) writes:
>
>a few weeks ago, i posted a request for information about
>non-holonomic motion planners.
>
>for those who are curious, i received some rather polite
>comments suggesting that
>
>(a) many people are interested
>(b) such planners do not exist
>(c) it would be nice to have such planners, but they are impossible
>(d) Jerome Barraquand and Jean-Claude Latombe have just the thing.
>
>while much credit for current work in this area should
>go to these two folks at Stanford, it's worth pointing out
>that there are at least three other groups working in this area too:
>- wilfong and fortune at Bell Labs
>- john canny and his students at Berkeley
>- jean paul laumond and his students at LAAS/CRNS, in Toulouse France

If your interest is car-like robots there is a very simple algorithm
developed independently by Francois Pin at Oak Ridge and myself.
Essentially, the algorithm is to go forward until blocked by an obstacle
turning in the direction of the goal, then back up turning in the
opposite direction.  As far as I know the details of the algorithm have
not been published, but I have a paper I can send to anyone who is
interested.  There is no formal analysis of this algorithm as far as I
know.

Erann Gat
gat@robotics.jpl.nasa.gov

