Newsgroups: comp.robotics
Path: brunix!news.Brown.EDU!noc.near.net!howland.reston.ans.net!darwin.sura.net!spool.mu.edu!caen!batcomputer!munnari.oz.au!cs.mu.OZ.AU!muse.cs.mu.OZ.AU!cdillon
From: cdillon@muse.cs.mu.OZ.AU (Craig Dillon)
Subject: Robot Simulator
Message-ID: <9315412.15120@mulga.cs.mu.OZ.AU>
Sender: news@cs.mu.OZ.AU
Organization: Computer Science, University of Melbourne, Australia
Date: Thu, 3 Jun 1993 02:48:34 GMT
Lines: 50

Hi All,

I have been receiving a constant stream of emails asking for further info
about the robot simulator that I was involved with writing, so I am 
reproducing that info here:

The simulator is written in C, has a GUI which runs in real time on a SGI
PI style machine, but can be configured to run in no-graphics mode on other
machines.

The simulator simulates the full kinematics and dynamics of a multi-link
robot, including full inertia, force, acceleration, velocity and position
calculations. The visual display draws the robot arm in nicely rendered
graphics (courtesy of SGI GL) at about 8-10 frames per second on a standard
PI machine.

Discrete time controller design is supported - each joint has a number of
possible sensors (position, velocity, acceleration and force) that you can
use as inputs (as well as a set of "desired position" values), and you set
which output you want to control on each joint. That is, if you are using
DC steppers, you control position. If you are using synchronous motors or
encapsulated stepper-controllers, you use velocity. If you are really
adventurous, you can simulate acceleration motors (if such things exist),
or force motors.

Graphs of joint velocities, accelerations, forces and positions can be
generated on the screen or into post-script format. There is a very crude
3D path planner, some demo robots, and a resonable set of documentation
on "how-to-use" as well as "how-it-works".

The simulator was written by Andrew Conway and myself, and is available
for distribution under the FSF copyright guidelines (which means you can copy
it for research/personal use, but not for selling - I don't think you
would want to anyway).

Our disclaimer is that this was in fact a 4th year undergraduate project, and
we only spent about three weeks doing it. Since then, my research interests
have swung more to vision, but Andrew is still involved in robotics, and is
now at Stanford Uni (arc@leland.Stanford.edu). 

You can get the source/docs from anonymous ftp on krang.vis.citri.edu.au
(note - this may change, email me if you have any troubles). It is under
/pub/robot.

If a reference is needed, email me and I will fix the docs up as being a
tech report (I just haven't gotten around to it...)

Craig Dillon
cdillon@cs.mu.oz.au

