SFU Computing Science Undergraduate Win Technical Achievement Award in Robotics Competition

January 18, 2008
Print

Written by Salima  Vastani

A team of intelligent and enthusiastic undergraduates from the School of Computing Science at Simon Fraser University (SFU) recently received the Technical Achievement Award at the Richard Tapia Robotics Competition that was held at the Richard Tapia Celebration of Diversity in Computing Conference in October 2007.

The winning team -  Nexus 6 -  made up of Lorin Beer (former sheet metal mechanic apprentice, and pursuing his Engineering Science degree), Angelina Fabbro (majoring in both Computing Science and Cognitive Science with a minor in Psychology), Angelica Lim  (recipient of The Canadian Federation of University Women and the Institute of Electrical & Electronics Engineers Award) and Kate Tsoukalas, (holds two B.Sc.’s , Computing Science and Applied Physics , and is currently in pursuit of her M.Sc. in Computing Science at SFU), won the coveted award. Their supervisor, Dr. Richard Vaughan, head of the Autonomy Lab at Simon Fraser University, also won recognition for being an excellent supervisor for the team.

The students, all from different backgrounds, held one similar interest – Robotics. It was part CS community and part coincidence that brought these four students with diverse yet relevant backgrounds to join hands for the Richard Tapia Competition.

The competition was a modeled search and rescue task, where autonomous robots had to navigate a maze and find unique coloured markers or survivors in configurations of either 2 or 3 colours and report them as found. The locations of the markers were not known to the participants before the competition actually started. The idea was to correctly identify as many markers as possible, as quickly as possible.

The iCreate robots that were to be used by the competing teams came with bumper sensors in the front of the device. This means that if the robot bumps against a wall or obstacle, it turns back and changes course. However, the Nexus 6 team programmed the robot in such a way that it does not bump into obstacles but changes course after identifying obstacles from a distance, thus saving precious time. Based on degree and range values returned by the laser range finder, each obstacle would create either a repelling or an attractive vector, depending on whether it detected an obstacle or a survivor object in the distance. The robot then either changed course away from the obstacle or moved in the direction of the survivor object.

The Nexus 6 team’s robot had one more unique attribute. It was programmed with two levels of curiosity. At the first level, the robot would explore the environment for a certain period of time, say 75 seconds. If it did not detect survivors, the robot would become more curious and explore more areas starting from the farthest detectable object. Even though the robot would go near a wall or any other object to detect if it was a survivor, it would rarely touch it. This way, the robot did an excellent job of avoiding obstacles compared to the other teams. Not only did it save them time but it was a lot more fluid while navigating through the maze.

The point of distinction that won the Nexus 6 team the prestigious award is quite probably this unique navigation algorithm they implemented. As team member, Fabbro said, “the highlight was the laser navigation. It doesn’t use any odometry. What the wheels are doing is almost irrelevant. It is the environment that is navigating the robot. This non-odometric solution to the navigation is probably what won us the award”.

“It was a navigation and picture-recognition challenge” stated Beer. “We encountered new and interesting challenges at every step” he said reminiscing about the 24 straight hours they put in to debug the laser just before leaving for the competition; this among the many technical challenges they faced up to the on-site competition.

When asked how they feel now about the entire competition and about robotics in general, the Nexus 6 team says “this feels like something we should be doing, this feels like something we should be involved in,” i.e. the Richard Tapia Conference for Diversity in Computing and Robotics. “We all want to get deeper into robotics, pursue a career, and go on to graduate studies. This seems like the path to be on in order to do that.” They also plan to “definitely go again to the conference.”

Pursuing their interest in robotics, these students are in the process of starting a Tea and Robotics Club at Simon Fraser University in order to attract more students to robotics.

Greg Baker, SFU Computing Science Undergraduate Program Director sums it up by saying "Nexus 6 and the innovation they used in their robot is an example of the creative thinking that we encourage our students to use when problem solving.  We are extremely proud of the Nexus 6 team, and look forward to their future robotic endeavours."