> home
> About
> Contact Us
> Editorial Info
> IEEE-USA

06.07

Backscatter: The Mouse That Wouldn't Quit

By Donald Christiansen

It all began in the early months of 1977. I was intrigued by the Machine Design magazine-sponsored competition, “The Great Clock Climbing Contest,” in which mechanical mice powered only by a battery and small electric motor were challenged to climb a vertical mesh “clock.” It occurred to me that real mice were noted for more than simply running up clocks, “Hickory Dickory Dock” notwithstanding. Their reputation as friendly, intelligent creatures who are particularly adept at finding the shortest path through a maze after first exploring various dead ends was well established.

Thus, in the May 1977 issue of IEEE Spectrum, I posed a challenge to its readers—then some 250,000 electrical and computer engineers. Why not design a maze-solving creature, with its own self-contained logic and memory—a “micromouse”—which could successfully navigate a maze which our engineer-editors would design, and the configuration of which would be held secret until racetime (a “mystery mouse maze”). The micromouse that demonstrated the best time would receive $1,000. Other prizes—for the best learning mouse, for example—would also be given.

And so was born The Amazing MicroMouse Maze Contest. By April of 1978, more than 5,700 eager entrants had signed on, over 300 from outside the United States.

But the challenge proved greater than most anticipated. The first time trials, in June 1978 at the National Computer Conference in Anaheim, drew only five contestants who had actually built and debugged their mice, and only two of the mice made it through the 8 x 8 foot maze.

The audience at the second time trials was enraptured, and included reporters from the Associated Press and The New York Times. But only one mouse made it through the maze. What an embarrassment! One newspaper headlined a piece by science writer Malcolm Browne “Mouse race big flop.”

Better days

Things picked up, as both designers and their mice grew smarter. Eight micromice ran the course at the third time trials at WESCON ’78 in Los Angeles, and fifteen competed in the finals at the 1979 National Computer Conference. The smartest mouse, designed by a team of engineers from Battelle Northwest Laboratories, cut its first run time of 1 minute 48 seconds to just 31.16 seconds in its third attempt. However, the $1,000 prize went to a semi-smart but very fast mouse with little decision-making ability. Also built by the Battelle team, that mouse’s strategy was based on a simple algorithm that would be useful to any child (or adult) who otherwise might panic inside an arboretum hedge maze. That is, you will escape the maze by simply feeling your way along one wall (either right or left). The winning micromouse was a “wall-hugger” that used light sensors to follow one wall, and a microprocessor to rapidly carry out the wall-hugging algorithm.

Outfoxing the wall-huggers

With due respect for the successful wall-hugging mice, we nevertheless wanted to encourage the design of smarter mice who could map the maze in early runs and avoid repeating missteps in subsequent runs. To discourage wall huggers, I designed a simple “mouse befuddler” that would cause them to loop in circles at the beginning of the maze.

The micromouse challenge caught on beyond expectations, becoming an international phenomenon. Micromouse mania soon spread to Europe and Japan. The first European competition took place in London in 1980. The following year another was held in Paris. The Boston Mouseathon, sponsored by the IEEE Computer Society, was held in 1985 at the Computer Museum. Entries from England and Japan participated at the Boston event. In 1985, the Japan Micromouse Association announced the first World Micromouse Content. The Institution of Electrical Engineers hosted the world competition in London in 1987, the same year in which the first Singapore Micromouse Contest was held. I was invited to judge the Singapore event. Sponsored by the Institute of Engineers of Singapore (IES), impressive entries came from Singapore Polytechnic, the University of Singapore, the French Singapore Institute, and the Japan Singapore Institute, among others.

In 1989, the IES sponsored the first International Micromouse Competition. In 1990, the North American Micromouse Championship was held at the Ontario Science Center. In 1991, the World Championship was held in Hong Kong, with 30 entries from thirteen countries. Among its organizers were the IEEE Hong Kong Section and the IEE Hong Kong Branch.

By the 1990s, the pace and number of micromouse contests had increased drastically. One online resource reported that “from five or six contests a year there are now over 100.” The May 1992 issue of the Mouser, a newsletter of the International Micromouse Community, alerted its readers to upcoming contests in London, New Zealand, Hong Kong and Australia. Micromouse clubs began to appear in engineering schools, and annual contests became regular fare. In the United States, the IEEE and the Computer Society sponsored many of them. In 2004, I was privileged to help judge the Micromouse Competition at the IEEE Region 1 Student Conference.

Micromouse Monikers

The designers proved to be as creative in naming their micromice as they were in designing and building them. It began with the Battelle team’s entries. Moonlight Express, Moonlight Special and Moonlight Flash were so named because they were built outside regular working hours. Harvey Wallbanger was a four-wheel wall-hugger and successful competitor in the IEEE Spectrum contest. Other early contenders were Theseus, Cattywampus, Mazey and Charlotte. Cattywampus was noted as “an ingenious but noisy specimen,” and won a prize for the most original looking mouse. Charlotte, “The Belle of Philadelphia,” featured long, seductive black eyelashes and an Intel 8748 microprocessor. Dudley and Mushka hailed from the University of Waterloo, Canada. Monty Mouse was “an aluminum sandwich on wheels” from Britain. His moment of fame came when he appeared on the British TV show “Science in Action.” Topo, a tiny mechanical wall hugger from Turin, Italy, crossed the Atlantic by himself and was met by a surrogate trainer for his shot at the first Spectrum time trials. A family of famous mice from MIT were appropriately named MITee Mouse I, MITee Mouse II, etc.

Tribulations

Not every contender found the going easy. The designers of Theseus reported that the mouse arose “out of endless Saturdays in the lab, numerous burned out semiconductor devices, mangled DIP chips, and shorted power supplies.” Monty Mouse’s designer reported that at its first run it “thought about where it was and then had a nervous breakdown and started smoking at the edges.” Fred, a micromouse from Plessey, gained notoriety for his unintended ability to spin in circles. Infrared sensors were mounted on the ears of Major Tom, but he courted disqualification at his first Wembley contest because of being mistaken for a tomcat. Thumper, a British mouse who participated in the Boston Mouseathon, was slow and lumbering, but talkative, with remarks such as “I will find the shortest route!” and “I hope there are no cats in here.” Though a winner in previous contests, Thumper lost out at the Mouseathon when, as one observer put it, he talked his way into a corner.

Press notices

Malcolm Browne’s putdown aside, the micromouse amassed an impressive scrapbook of press clippings. The finals of the Amazing Micromouse Maze Contest were covered by CBS, NBC and ABC television and reported by Walter Cronkite and John Chancellor and David Brinkley on their evening news shows. Moonlight Special appeared on the front page of The Wall Street Journal. Science magazine headlined its story “Microcomputer: The Great Electronic Mouse Race.” The Salt Lake Tribune opened with “Smart Mice Will Beat a Path to this Contest Door.” US magazine said “Move over Mickey. R2D2 is here.” A broad range of newspapers—from the International Herald Tribune to the Boonville (Mo.) News reported on the event.

The Squeak Goes On

Thirty years from its inception, the micromouse appears to be going strong. Among countless 2007 micromice challenges and demonstrations are the following: the UK National Micromouse Competition, the Singapore Inter-school Micromouse Competition, and the All-Japan Micromouse Contest. In January, the International Micromouse Competition was held at Techfest 2007 in Bombay. The Applied Power Electronics conference (APEC 2000) held its 21st Annual Micromouse Contest at Disneyland in February. UK Micromouse 2007, scheduled for 30 June at Millennium Point, Birmingham, England, features separate heats for contact wall followers, non-contact wall followers, and maze solvers. It includes a separate beginners’ contest, Micromouse for Schools, aimed at teams whose members are 18 or younger. The school teams’ mice are challenged to follow a line or a wall but are not introduced into the maze itself.

Among the IEEE Student Conferences at which micromouse events are scheduled are those of Regions 1, 2 and 6. The University of California at Davis held a Picnic Day Micromouse contest in April. Santa Clara University planned a National Engineers Week micromouse demonstration and geared up for its second try at the Region 6 conference. A micromouse competition is planned for Robothon 2007, Seattle, Wash., 21-23 September.

The Bottom Line

Aside from its being a fun, competitive challenge, the micromouse has proved to be an excellent teaching medium. It can be viewed as a small system involving interdisciplinary engineering aspects. Its successful designers often work in teams, and must consider electrical, electronic, mechanical and computer issues. Design decisions and tradeoffs involve weight, speed, power, sensing techniques, turning methods, center of gravity and programming.

Among the non-winners of micromouse contests, many have praised the challenge while cheering their competitors on to victory. One contestant who experienced the agony of defeat said of his design team “We learned much by having to design, create and try.” On a personal note he said the micromouse experience propelled his career forward.

Resources

For more about micromice and mazes, see:

• Christiansen, D., “Announcing the Amazing MicroMouse Maze Contest,” IEEE
  Spectrum, May 1977.

• Allan, R., “The Amazing Micromice: See How They Won,” IEEE Spectrum,
  September 1979.

• Allan, R., “Three Amazing Micromice: Hitherto Undisclosed Details,” IEEE Spectrum, November 1978.

• Matuszek, D., “How to Build a Maze,” Byte, December 1981.

• Billingsley, J., “Micromouse: Mighty Mice Battle in Europe,” Practical
  Computing, December 1982.

• Billingsley, J., “Micromouse: Maze Mastery,” Practical Computing, September 1983.

• Braunl, T., “Research Relevance of Mobile Robot Competitions,” IEEE Robotics and Automation Magazine, December 1999.

• Braunl, T., Embedded Robotics: Mobile Robot Design and Applications with Embedded Systems, Springer-Verlag, 2003.

• Among those who warrant special recognition for helping expand interest in
  micromouse design are David Otten of MIT MITee Mouse fame and John
  Billingsley, who fostered interest in Europe.

Donald Christiansen is the former editor and publisher of IEEE Spectrum and an independent publishing consultant. He can be reached at donchristiansen@ieee.org.