RI: Small: Robotic Search of Transient Objects
- View All
Searching for objects in physical space is one of the most common tasks for robots. Transient targets refer to a class of objects which are not identifiable unless momentary sensing and/or signaling conditions are satisfied. The transient property is often introduced by target attributes, privacy concerns, environment constraints, and/or sensing limitations. For example, searching for a black box on the ocean floor after an airplane crash is a typical transient target search (TTS) task because the search relies on the transient radio/sonar signals emitted from the black box. This project develops the theoretical foundations for TTS problems that will impact a large group of real world applications. Due to their stochastic nature, TTS problems are challenging because the transient property is often coupled with factors such as sensing range limits, various coverage functions, constrained mobility, signal correspondence, limited number of searchers, and a vast searching region. Extensive modeling and analysis will be performed to understand the various factors in the searching process. As the result, a new computation framework consisting of models, metrics, and algorithms that integrate the resource to improve TTS performance will be developed with three focuses: searching time analysis that addresses the critical performance issue, multi-target search that focuses on unique signal correspondence and collaborative sensing issues, and planning for coordination of robots which includes both centralized and decentralized approaches. This project also develops new course materials for undergrad and graduate students to better understand algorithms and robust intelligence in robotic searching applications.