Enabling Unmanned Aircraft Systems to Detect and Avoid Other Aircraft

August 2013
Topics: Unmanned Systems, Civil Aviation Security
MITRE is exploring concepts and technologies to support the safe integration of unmanned aircraft into the nation’s airspace system.
uav pilots

Unmanned aircraft systems (UAS) have proven invaluable in military operations overseas and in homeland security and public research missions domestically. The Federal Aviation Administration, the National Aeronautics and Space Administration (NASA), the Department of Defense, and the Department of Homeland Security seek to routinely integrate these aircraft into non-segregated civil airspace for uses such as public safety, scientific research, weather monitoring, border patrol, and eventually commercial operations.

One mutual concern revolves around UAS remote pilots' inability to "see and avoid" other traffic in shared airspace. This is especially challenging when they try to maintain safe separation from other traffic flying under Visual Flight Rules, which may not be transponding in response to secondary surveillance radars.

Ground-Based Sense and Avoid Support

A promising option is the Ground-Based Sense and Avoid (GBSAA) concept. With GBSAA, a radar system provides the remote pilot with traffic information for the operating airspace. This enables the pilot to maintain safe separation between the UAS and other aircraft. Emerging research associated with manned aircraft flight has demonstrated that cockpit displays of this traffic information might be effective in enabling aircraft separation and preventing potential conflicts.

Automated Detect and Avoid Research

MITRE is spearheading a multi-organization research effort—the Limited Deployment-Cooperative Airspace Project (LD-CAP)—to explore the viability of a capability for an unmanned aircraft to automatically detect and avoid conflicts with potential intruder aircraft, without any input from a pilot required.

In conjunction with the University of North Dakota (UND), NASA, Draper Laboratory, and other institutions, MITRE is exploring the operational feasibility of automatic detect and avoid algorithms, or computer programs, that enable an unmanned aircraft to automatically maneuver based on the data it receives from a sensor providing information about other aircrafts' locations and movements. The LD-CAP research is employing a sensor called Automatic Dependent Surveillance-Broadcast (ADS-B).

All aircraft equipped with ADS-B are broadcasting their positions to one another on a regular basis (approximately once a second). With that in mind, MITRE researchers are exploring the possibility of using ADS-B surveillance capabilities, combined with computer algorithms, to allow UASs to automatically maneuver to avoid conflict with other aircraft.

After simulating millions of encounters in MITRE's algorithmEvaluator, the LD-CAP team took the testing from the computer into the sky to both evaluate the operational feasibility of such algorithms and to validate the computer-based simulations. The team conducted a series of three separate flight test campaigns using automatic research algorithms developed by MITRE, UND, and Draper.

These tests occurred in July and September 2013 in airspace in Eastern North Dakota and in August 2013 in airspace near NASA Langley in Hampton, Virginia. Based upon the data collected from hundreds of real-world encounters using a variety of encounter geometries compared to data from millions of simulated encounters, the team believes the computer-based simulations show reasonable congruence to actual flight and thus would have significant utility in evaluating the performance of prototype automatic detect and avoid algorithms.

NASA modified a Cirrus SR-22 to operate as a surrogate UAS with the ability to be controlled by a general purpose computer on-board the aircraft. A variety of automatic algorithms could be loaded on the general purpose computer, which is linked to an ADS-B In transceiver. A variety of intruder aircraft were operated, including aircraft provided by UND, NASA, and MITRE using a number of different encounter geometries. In addition to ADS-B Out, the team also evaluated the performance of ADS-R, TIS-B, and the potential for a TIS-B message based upon primary FAA radar feeds as the basis of tracking an intruder. At all times, a NASA safety pilot was in the cockpit of the surrogate UAS. The automatic algorithms all were able to automatically maneuver the surrogate UAS to avoid conflicts with the intruders.


Publication Search