ABSTRACT

Wilderness search and rescue (WSAR) is the task of finding and giving assistance to humans who are lost or injured in mountain, desert, lake, river, or other remote settings. Because of the vast distances involved in wilderness settings, searchers frequently depend on surveillance from helicopters and small airplanes. Although these resources are very useful for searchers, the have limitations: resources consume considerable cost, there can be delays between when the resources are needed and when they arrive, ground searchers and pilots must overcome communications barriers between them, and the aircraft may not be able to provide low level imagery because of flying restrictions associated with rugged terrain. The central hypothesis of this project is that mini (3-5 foot wing spans), fixed wing Unmanned Aerial Vehicles (UAVs) can be used by WSAR personnel to efficiently find people in the wilderness. The human factors issues associated with small UAVs are much different than those associated with large UAVs, mostly because small UAVs for WSAR personnel imply limitations on operator training, sensor capacity, autonomy capability, and flight time. The PI's plan is to develop operator interfaces and UAV autonomy for WSAR systems that allow people without RC-piloting skills to search an area, using either online or offline approaches. When working online, the PI will adopt a non-pilot operator perspective and design autonomy to allow operators working in an "augmented virtuality" environment to "guide the camera" rather than fly the UAV. In situations where information from a UAV's video is to be recorded and used in offline information retrieval and analysis, the PI will pursue an active mosaic approach in which video images are overlaid on terrain maps. The PI will employ a strongly human-centered approach in all phases of the project, both for creating the WSAR systems and for evaluating them, in which expertise from researchers in human-robot interaction, computer vision, controls, and artificial intelligence is integrated. User studies will include field tests with WSAR personnel, investigation of current work practice in WSAR teams, usefulness of active mosaicing for offline and online searches, and so on.

Broader Impacts: Each year, many people are lost or find themselves in jeopardy while hiking, boating/kayaking, skiing, fishing, etc. Each year, wilderness search and rescue consumes thousands of person-hours and hundreds of thousands of dollars in Utah alone. With each hour that passes between the time that a person is lost and WSAR people find the victim, the effective search radius grows by approximately 3km. Each hour spent in the water or lost in the woods decreases the likelihood of a successful rescue. A portable UAV with appropriate interfaces, autonomy, and sensor processing at an affordable price should decrease the amount of time required between when searchers arrive at a scene and the time when aerial surveillance is present to support their efforts. Such a system would increase the probability of successful rescue.

Please report errors in award information by writing to: awardsearch@nsf.gov.