Unmanned Aerial Systems (UAS) have gained widespread popularity for several applications including Intelligence, Surveillance and Reconnaissance (ISR), border patrol, disaster relief. However, practical issues in sharing the national airspace (NAS) safely with other manned aircraft has been a significant challenge. The main technology barrier is that the UAS lack the ability to "Sense and Avoid" other intruder aircraft in the NAS. The SAA capability will allow UAS to achieve an "equivalent level of safety", comparable to the "see-and-avoid" requirements for manned aircraft and ensure that UAS do not disrupt other airspace users.
In general, SAA can be decomposed into three main subfunctions: detecting and tracking intruder aircraft using onboard sensors, performing collision detection, i.e., deciding whether the intruder aircraft is on a collision course, and avoiding potential collisions. At UtopiaCompression, the focus is on developing low cost, size, weight and power (CSWaP) SAA solutions for small and medium-sized UAS. I am involved in several SAA projects from the Air Force and DARPA that are focused on the aforementioned three aspects of SAA.
Developing a Laser RADAR (LADAR) for SAA
We have designed a LADAR system to enhance existing passive systems for SAA applications, such as target tracking, ranging, and false alarm rejection.
Maneuverless Monocular Passive Ranging System (M2PARS)
We have developed a passive ranging algorithm that estimates range of moving targets without ownship maneuvers using only monocular cameras!
Collision Detection and Avoidance with Passive Sensors
Leveraging the range estimates from M2PARS, we have developed an effective collision detection rule that classifies oncoming air traffic into collision threats and non-collision traffic.
Work with UtopiaCompression team featured in the news
http://www.darpa.mil/news-events/2016-04-05a
http://www.engadget.com/2016/04/06/darpa-uav-mid-air-collisions/;
http://unmanned-aerial.com/darpa-flight-tests-demonstrate-optical-sense-and-avoid-capability/
Current work: Maneuvering intruder passive ranging
Publications
[C27] J. Daugherty, H. Bai and S. Avadhanam. Orbiting Intruder Passive Ranging for Small UAS Detect-and-Avoid. In the Proceedings of the first IEEE conference on Control and Applications, 2017.
[P1]. A. Ganguli, H. Bai, S. Avadhanam, J. Newton, A. Tiwari, and J. Yadegar. Maneuverless Passive Range Estimation using Monocular Image Sequences, Patent No. 9,342,746, May 17, 2016.
[C18] H. Bai, et al. Developing a vision-based sense and avoid system. In the Proceedings of the AUVSI Unmanned Systems conference, Atlanta, GA, 2015.
[C14]. L. Sahawneh, R. W. Beard, S. Avadhanam, and H. Bai. Chain-based collision avoidance for UAS sense-and-avoid systems. Accepted by AIAA Guidance, Navigation, and Control Conference, Boston, MA, 2013.