Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!hsdndev!husc6!jbaker From: jbaker@husc9.harvard.edu (James Baker) Newsgroups: comp.robotics Subject: constraint programming and robot motion Message-ID: Date: 16 Jan 91 16:01:01 GMT Sender: news@husc6.harvard.edu Organization: Harvard Arts and Sciences Computer Services, Cambridge, MA Lines: 26 I am currently working on a new programming language for describing explicit motion. Although I realize that there are zillions of robot languages out there, I believe this one is somewhat unique in that it is founded on a formal relationship (which I am still developing) between Matt Mason's C-surface model for compliant motion and constraint programming. The base constraint satisfaction system is Wm Leler's Bertrand. (See Leler, Constraint Programming Systems, Addison-Wesley, 1988.) To go into more detail, CLEAR (Constraint Language for Environment And Robot) is to employ and preserve the nonprocedural (read parallel) semantics of constraint programming while giving the programmer a means of describing a procedure (without say an explicit reference to time, which is clumsy and nonintuitive especially for quasistatic operations). CLEAR also endeavors to be object-oriented but in a subtly different manner than more conventional object-oriented languages. One advantage of this object orientation, incidentally, is that it should facilitate compilation of the high-level description seen in constraint programming to something like C or C++. I would be interested in any thoughts you might have. Additionally, if anyone has seen a good robot simulation system on UNIX that is free or cheap and is easy to interface with, please write to me. -- Jim