DPRG
DPRG Events Shoptalk

Website design and hosting by NCC

GARP (Group Autonomous Robot Project)

March 1997, Original project summary by Jim Brown
August 2002, Massively revised and reformatted by R. Steven Rainwater
December 2005, Minor revisions and updates by R. Steven Rainwater

The Origin of GARP

GARP Wheelchair March 1997

By early 1997, the idea of building a DPRG group robot had come up several times in DPRG meetings. Lots of members liked the idea but no one ever seemed to have the perfect plan or the motivation to get it started. In February of 1997, Jim Brown, then the president of the DPRG, communicated with someone working in a medical supply store somewhere in New Jersey. They had an old, powered wheelchair that they'd like to anonomously donate to the DPRG.

The DPRG paid $41.95 for the shipping and, by March 1997, the wheelchair had arrived. It was brought to the March DPRG meeting. After some time spent deciding on a name for the project, Larry Kerns suggestion of Group Autonomous Robot Project or GARP was selected. There was an initial flurry of activity over the next several months as GARP began to take form.

GARP Unveiled

GARP April 1997

An initial inspection of the wheelchair revealed that it was not in working order. The motors appeared to still be functional but there was a lot of wear and some damaged wiring. James Vroman volunteered to take it home and do a throrough checkout.

When GARP was next seen at the April DPRG meeting, James had cleaned and reconditioned much of the mechanical and electrical system including replacing wiring. A little reverse engineering produced a schematic of the generous volume of wiring to be found in the seemingly simple system. James constructed a wooden case to hold electronics and computer hardware that fit onto the top of wheelchar base without the need to alter the structure (thus allowing it to be removed should GARP wish to return to duty as a wheelchair).

GARP Evolves

GARP April 1997

Using primarily parts from the infamous Tech Tools dumpster, James fabricated a simple two-speed tethered controller and demonstrated it by driving GARP around the room. A lot of donations were received from other group members for the next stage of work. Among them were a 40MB hard drive and TTL monitor donated by Dan Mathias, two 386 motherboards from Roger Arrick, a 630MB SCSI drive from James Huynh, a 12v boat battery from James Vroman, a VGA monitor from Brian Merritt, a VGA card from Eric Yundt, and a mini-tower case from Jim Brown. There was much talk of running the Linux operating system on the motherboard communicating with either 8051 or HC11 processors to run actuators and sensors.

James Vroman continued to work on GARP throughout 1997. He planned on creating an open architecture for the robot that would allow other group members to easily build components to interface with it. He also began adding Molex connectors to all control circuitry, motors, and the power system to allow alternate types of components to be swapped in and out at will. A 12V and 5V power bus were planned as were RS-485, RS-232, and a simple ISA bus IO board for the 386 motherboard. Jim Brown created a web page write up of the project (which eventually evolved into what you are now reading).

GARP and Punctuated Evolution

Work continued at a slower and slower pace through most of 1997. In part the slowdown was due to James Vroman moving to Missouri. He had been the prime mover on the project and, without him, the project lost focus. At some point, the wooden case built by James was removed, the top of the wheelchair frame was cut off, and a flat, laminate surface was mounted on it. The intention was now to permanently mount control hardware on the flat surface but interest began to wane again at this point and by the end of 1999, the project had gone into stasis.

From time to time after that, someone would bring GARP up at a meeting, a brief flurry of activity would follow, and then a slow return to stasis occured. The fundamental problem seemed to be that no one was really interested enough in having a wheelchair-based robot to devote the time needed. Several people tried to jumpstart the project again, including Bob Jordon. Bob donated some heavy duty H-Bridges to be used for motor control. Ed Okerson later did some work connecting the H-Bridges and getting GARP back to the point where it could move under its own power. Ed also reverse engineered the joystick and control box wiring. He had plans to add a motherboard running Real Time Linux to control GARP. Things were briefly looking hopeful again when Ed decided to move to California.

The Curse of GARP

From that time on, GARP resided in an unused bathroom at the DPRG warehouse. The rubber components did not age well and the tires became permanently flat and began decaying. There was some talk in 2002 of working on a DPRG Group Robot and the subject has come up from time to time since then. GARP is usually brought up whenever such talk ensues, more as a warning or a proof that group projects are a waste of time than anything else. Sadly GARP turned into something of a curse on group projects in the end. Several attempts have been made in recent years to dispose of GARP but, like the cat in the well known folk song, GARP always seems to return to the warehouse. Current theory holds that the only way we'll finally get rid of GARP is to encase it in cement and drop it into the Pacific Ocean over the Mariana Trench. One member proposed that it may be necessary to transport the rusting hulk of GARP across the shadowy land of Mordor and cast it into the fiery chasm of Mount Doom in order to break the curse.

The Details

Copyright © 1984 - 2014 Dallas Personal Robotics Group. All rights reserved.