Robots Vie for Best Balance at APL
Team five, the undisputed winners of the Lego Robot Competition, were able to balance their robot for an impressive 26 seconds
On a bitter cold evening in early December, five student teams gathered in a computer lab in the Applied Physics Laboratory’s Kossiakoff Center. Each group of four to five students spent approximately 45 minutes fine-tuning their handiwork: identical, 8-inch-high LEGO robots. The teams were preparing their Segway-like robots for an annual competition, the culmination of a semester’s worth of learning in the Johns Hopkins Engineering Programs for Professionals (EP) computer science course, Software Development for Real-Time Systems.
In the master’s level course, students examined the hardware and software technologies behind real-time, embedded computer systems. Embedded computers are utilized in many modern devices, from smart kitchen appliances to sophisticated flight control systems for airliners and yes, robots.
“These students are already software and electrical engineers,” said Jeff Gustin, an instructor in EP’s computer science program. “The competition helps them reinforce concepts and basic principles. I selected the LEGO Mindstorm robots because you can do some sophisticated programming with them.”
For Gustin’s graduate students, the spirit of the competition itself hones skills that will be valuable in the workplace since students must solve real-world problems and work as teams under pressure.
The students had only three class periods – approximately three hours – to construct and program their robots. Using the Java programming language, students wrote and then downloaded a balancing program into their robots. Given the limited timeframe, the student teams didn’t have a chance to give their robots names, but they all seemed attached to their creations nonetheless.
The Competition Begins
Just before 8 p.m., the student teams gathered one by one in a large open space in the Kossiakoff Center. In preparation for the competition, Gustin placed a large piece of white posterboard on a table.
Competition rules were simple: each team could try five separate times over a period of five minutes to balance their robots on the posterboard. Each team selected a spokesperson who could briefly discuss challenges the team faced and how – and if – they overcame those challenges. The “winner” was the team whose robot stayed balanced the longest.
Team 4 was the first competitor. The team, which had programmed a sleep function in the robot for stability, took measurements of the light value to calibrate their creation. They also switched the battery in their robot early on, but their longest time was just over one second as their robot looked as if it would balance, then zipped backward off the posterboard into the waiting hands of a “spotter.”
The Team 1 robot was next. Team 1, looking for “someone with steady hands,” used game coefficients in its robot’s programming and acknowledged that they may have corrected too much. The group also felt the robot’s light sensor was too close to the paper. The spokesperson said, “We broke the first rule of making a presentation – we wrote new code.” As the Team 1 robot swayed back and forth before dashing for the edge of the posterboard, a team member asked, “Are we giving up?” The answer: “Yes.”
Team 3 was the third group to compete, and they had built a small rig to help give their robot more balance than they felt hands alone could do. After several unsuccessful attempts, the team thought their light sensor was too high and tried moving the sensor closer to the paper, but to no avail. On the fourth try, Team 3 unveiled its special rig and after balancing for a second, their robot reared back and zoomed off the table.
Up next, Team 5 used a pre-selected team member as the “official balancer,” and their robot stayed in balance for an impressive 26 seconds. The team’s secret weapon might have been a timer that went off every six milliseconds to recalibrate the robot and maintain balance.
Team 2 was the fifth and final group to compete. The spokesperson admitted they had used “unorthodox approaches,” and the team tried changing the robot’s power settings in an attempt to achieve balance. As their robot sped for the posterboard’s edge, one team member remarked, “I think what really happened is someone switched the stickers before we came down here.”
Team 5 was the undisputed winner of the competition, with students James Ahn, Kevin Fitch, Jared Milburn, and Alex Watson comprising the team. Fitch said he had enrolled in the master’s program in computer science to further his career. A software engineer for Intellisys, Fitch believed the robot competition helped enhance his skills at working with his peers. “This required more teamwork than your average day at work,” he said.