The Tech - Online EditionMIT's oldest and largest
newspaper & the first
newspaper published
on the web
Boston Weather: 33.0°F | A Few Clouds and Breezy

‘Bots in Blue’ Complete Qualifying Round of 6.270

By Kevin R. Lang

Hackers will be jailed. Students will remain on either East Campus or West Campus. Professors will be escorted across Massachusetts Avenue.

Have the Campus Police resolved their labor dispute and expanded their duties at MIT? Hardly. This year’s 6.270 Autonomous Robot Design Competition, “Bots in Blue,” features robots trying to capture “hackers,” keep “students” on either side of “campus,” and escort “professors” across “Mass. Ave.”

Round one finished up Tuesday afternoon, with 27 of 60 teams qualifying for the final competition on the first try. The final contest begins Thursday at 6:00 p.m. in Room 26-100.

The first round featured false-starts galore, non-functioning robots, and some flying Legos. Many teams had been finishing work both on code and on their robots late Monday night and into Tuesday morning. Controller boards this year were significantly delayed, despite course organizers’ attempts to debug the boards since last year.

“We were just figuring out how long I’d been awake,” said contestant Casey R. Muller ’02.

First run robots yield low scores

Despite the potential to score several dozen points, the high score for round one was only twelve points. Many teams qualified with only one or two points; teams need only demonstrate the ability to score in order to qualify, regardless of winning or losing a given matchup.

Course organizers said that many teams typically need several chances to qualify, although some contestants argued that the controller board issues gave them less time to test and debug.

A number of teams did not attend the first round, several teams were disqualified after double false-starts, and still more did not qualify because their robots would not stop running when time expired.

In one match, a robot scored several points while pushing blocks into the other team’s goal. However, the opposing robot did not function. Thus, the non-functioning robot won the round without qualifying, but the losing robot qualified. Such odd matchups were not uncommon in Round One. In several matches, the placebo robot, employed during single robot matches, was victorious in much the same way. The relatively small crowed cheered the placebo robot more than any team’s creation.

Most teams are using similar designs this year, either pushing or pulling several hacker blocks into the jail. Only one team this year attempted to use a second, remote robot, although they did not qualify in the first round. Several teams are using extended Lego arms to sweep blocks across; one team scored twelve points by collecting all four professors in one sweep.

Teams use vastly different boards

The two boards being used by contestants are considerably different: the original runs Java in 16 MB of memory, while the second, more reliable board runs C in 32 KB of memory. “Both are more than capable,” said course organizer Anthony Y. Hui ’99.

One team that originally began programming for Java and switched to the C board thought the slower board was sufficient. On either board, “full speed is way too fast,” Muller said. “I really think that C is more powerful for this kind of thing.”

Muller and teammates Alea C. Teeters ’99 and Audrey L. Snyder ’03 qualified using a “scoop and sort” strategy to gather blocks, then sort them by color.

After round one, the 6.270 lab in Building 34 was nearly empty, despite the fact that 33 teams still needed to qualify. Hui thought most teams were recovering from late-night programming.

“I think they’re catching up on sleep and they’ll be back at like three in the morning,” Hui said. “Lab’s open twenty-four hours a day these days.”

Teams score points in this year’s contest by putting black blocks (hackers) into a “jail,” moving white blocks (students or non-hackers) into their end of campus, and pulling orange blocks (professors) out of the middle of “Mass. Ave.”