MARS by Jason L. Verhagen

MARS (Mobile Autonomous Robotic System) is an experimental robotic 
platform being designed with expansion, modularity, and
flexibility in mind.  The initial Phase 1 testing is being done with 
a small tracked vehicle.  Using this test vehicle I am
testing various sensors and programming code. 

The Phase 1 test vehicle can be controlled via IR remote, preprogrammed 
with a specific path, or roam unattended when
outfitted with proximity detectors.  The test vehicle can also sense 
stairs, light, and heat.

- Powered by 2 9.6V 1000mAh battery packs.
- Primary controller is a Handyboard with HB Expansion board.
- Two hacked 100 oz-in RC servos providing torque to tracks.
- One Basic Stamp II controlling 8 proximity modules.
- Four IR prox modules for contact detection (2 front, 2 rear)
- Four IR prox modules for stairs detection (1 at each corner).
- Three CdS cells to detect light (left, right, and up)
- Two thermistor temp sensors (1 external, 1 internal)
- One cooling fan (controlled by internal thermistor)

Parts used:
        The Electronics and Motors: 
1 x MIT Handyboard with Charger Board, LCD, and Expansion Board
1 x Basic Stamp II with Board of Education 
4 x GP2D05 IR proximity modules 
4 x IR LED/Detector pairs for stairs detectors 
3 x Radio Shack CdS photo cells 
2 x Radio Shack Thermistors
1 x Mini 12v cooling fan
2 x Dual connector phone jacks (for serial interfaces)
4 x Radio Shack terminal strips 

        The Chassis: 
1 x Radio Shack Bedlam (hacked for tracked chassis)
3 x Various Radio Shack project boxes
1 x Old external modem top cover
2 x Small pieces of cable shield tubing
various strips of brass stock (servo and battery mounts)
many cable ties
many machine screws and nuts

Problems Encountered:
- The motors that came with the Bedlam RC car were too powerful 
for my needs and I couldn't get them under control with an
LMD18200 motor controller I put together.  I replaced the motors with servos.
- The servos didn't give me the speed I wanted with the original 
Bedlam gears near a 1 to 1 ratio.  I put a bigger gear on the
output shaft of the servos to give approx. a 2 to 1 ratio.
- The increasing weight of the robot makes turning on it's axis 
difficult with the current 100 oz-in servos.
- IR contact sensors won't detect objects lower than the bumper in 
many instances - contact whiskers are going to be
implemented to catch lower objects.

Plans for Expansion:
- Build onto a bigger chassis (I am thinking about using a Power Wheels kids powered car)
- Using a laptop as a primary onboard computer
- Sonar with mapping and path planning on the primary computer
- Wireless video and audio feed from the robot
- Wireless data link to/from the robot
- Frame grabber on the robot for vision recognition
- Voice command recognition
- Whatever else I can think of :)

Time to build:
Approx. 50 hours over 2 months

Parts listed above: Approx. $800US
Tools and supplies: Approx. $350US

Website URL:

This site will host all of the details, schematics, and 
programming code that I develop.  Be patient, I work a full time (50+
hours) in a paper converting mill and I don't always have 
alot of time to work on the MARS project and getting my ideas and
information on the web site as quickly as I would like.  
However, feel free to email me with comments, questions, or
suggestions and I will reply as soon as humanly possible.  

I am currently developing the MARS project in an open source/gnu 
fashion so that others can freely study and learn from it. I
would appreciate the opportunity to collaborate with others 
via email about ideas for the MARS project or autonomous robotics
in general. 


Return to The Robot Menu