Sense-and-avoid (SAA) is a critical research topic for enabling the operation of Unmanned Aircraft Systems (UAS) in civilian airspace. SAA involves two planning related problems: 1) plan-recognition to predict the future trajectory of nearby aircraft, and 2) path planning to avoid conflicts with nearby aircraft that pose a threat. We have designed and built components of a novel intelligent sense-and-avoid (iSAA) reasoning framework that takes into account information about aircraft type, transponder code, communications, local routes, airports, airspace, terrain, and weather to more accurately predict near- and medium-term trajectories of nearby aircraft. By using this additional information both the onboard control software and the ground-based UAS operator can make more informed, intelligent decisions to effectively predict and avoid conflicts and maintain separation. While this capability benefits all categories of UASs operating under both Instrument Flight Rules (IFR) and Visual Flight Rules (VFR), it is absolutely essential for allowing smaller UASs to operate VFR at low altitude in uncontrolled airspace for operations such as survey work, wildlife tracking, aerial photography, utilities inspection, crop dusting, and package delivery.
[1]
M Berry,et al.
Literature Review on Detect, Sense, and Avoid Technology for Unmanned Aircraft Systems
,
2009
.
[2]
Chad Salisbury,et al.
Joint Airspace Management and Deconfliction (JASMAD)
,
2009
.
[3]
Hector Geffner,et al.
Probabilistic Plan Recognition Using Off-the-Shelf Classical Planners
,
2010,
AAAI.
[4]
Douglas Marshall,et al.
Introduction to Unmanned Aircraft Systems
,
2011
.
[5]
Mykel J. Kochenderfer,et al.
Next-Generation Airborne Collision Avoidance System
,
2012
.