A Large-Space Indoor Drone Testbed for Simulation of Search-and-Rescue Tasks in Complex Urban Terrain
Abstract
Among many military and civilian applications that concern autonomous or semi-autonomous robots, search-and-rescue tasks aimed for very complex urban terrain remain a high-priority, high-reward challenge for multiple agencies in the Department of Defense. Through an ONR MURI and its continuation programs, the Berkeley team has developed a new theoretical framework to achieve autonomous drone safety control and effective task supervision that outperform both existing autonomous solutions and human manual performance. Due to the limitation of space and safety regulations, our past experiments were conducted primarily in small indoor spaces, which presents a technology gap to validate and further develop our theories and algorithms for real-world outdoor scenarios. In this proposal, we have secured a valuable new large-space indoor facility, which would enable us to conveniently test fly large drones built for outdoor use in a lab environment just a few blocks from our current facility in the EECS Department. We plan to deploy and test our state-of-the-art drone safety planning policies on the much larger-scale drones, install and prototype more sophisticated onboard sensors for real-time urban terrain mapping and a real-world search-and-rescue scenario, and conduct extensive rigorous user interface and user experience studies to allow human users (especially war fighters in urban combat scenarios) to safely and effectively control autonomous drones and timely analyze fusion of multi-modal sensor data using new immersive 3D interfaces such as virtual reality and augmented reality.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 31, 2020
- Source ID
- N000142012729
Entities
People
- S. Sastry
Organizations
- Office of Naval Research
- United States Navy
- University of California Regents