Stafford, Kenneth A
Great strides have been made in improving the decision making capabilities of robots in unpredictable situations, but relatively little attention has been paid to the ability of robots to reason about--and compete with--human and non-human opponents. Such systems must collect and process data on the changing environment, anticipate actions of opponents, and navigate in often cluttered environments.
In this report, we present approaches for four different high-level tasks applicable to small, fully autonomous robotics platforms competing in a college-level extracurricular robotics competition. The goals of the project are real-time motion planning and maneuvering, long-term localization, object detection and tracking in a cluttered environment, and adversarial strategic planning.
Worcester Polytechnic Institute
Major Qualifying Project
All authors have granted to WPI a nonexclusive royalty-free license to distribute copies of the work, subject to other agreements. Copyright is held by the author or authors, with all rights reserved, unless otherwise noted.