Engineering students on the Robot Study Team (RST) at Trinity College in Hartford, Conn., took full advantage of a Thales Navigation BR2G-S global positioning system (GPS) receiver to finish among the leaders in a field of 25 university teams in the recent Intelligent Ground Vehicle Competition sponsored by the Association for Unmanned Vehicle Systems Intenational (AUVSI).
The IGVC competition is an annual event sponsored by the Association for Unmanned Vehicle Systems International (AUVSI), was held in July in Orlando.
The Thales receiver was mounted on the Trinity team's ALVIN III robot and accurately directed the autonomous vehicle to a series of checkpoints in the competition's Navigation Challenge, in which the vehicles were required to avoid various course obstacles in locating the points.
"The GPS receiver did its job to perfection," said Michelle Bovard, the Trinity team member who researched and selected the Thales receiver for ALVIN III. "The system consistently brought ALVIN to well within the required two-meter distance from waypoints provided for in the rules. ALVIN undoubtedly would have finished higher than the eighth place it earned if not for some unexpected problems our vision system encountered with some of the course obstacles."
ALVIN III is a 99-pound vehicle that uses its vision system to drive within lanes and avoid potholes, traps, barricades and other vehicles. The vehicle's sophisticated onboard vision system allows it to see obstacles in advance and plan a path around them. Alvin uses a differential drive system, consisting of 16-inch bicycle wheels powered by electric motors.
The Thales Navigation receiver was selected for its combination of accuracy, ease of use and light weight, according to Bovard. "We needed a differential system with sub-meter accuracy that would be relatively easy to work with and relatively lightweight. The Thales system was a perfect fit," she said.
First step: incorporating GPS data with software
Once the Thales BR2G-S unit had been selected, the team worked to incorporate GPS data into its software. "The receiver was easy to use, and on the few occasions we needed assistance, the help lines were great," she said.
The team used Coast Guard Beacon signals to get DGPS corrections, both in their Connecticut practice sessions and in the Florida competition. "We had little trouble getting the satellite and DGBS corrections, especially in Florida, where we were within range of three beacon stations," Bovard said. "The GPS consistently gave us good data and got the robot to within .5 meters of our practice waypoints."
The Trinity team got a late start this year in preparing for the Navigation Challenge, though Bovard indicated that won't happen in next year's competition. "We already have plans for more complex navigation algorithms and improved integration of the navigation and vision systems," Bovard said. "We plan to start work as soon as we get back to school in the fall. We have high expectations for next year."