Detect and Avoid Alerting Performance with Limited Surveillance Volume for Non-Cooperative Aircraft

Abstract

This paper investigates effects of limited surveillance volume on the alerting performance of a Detect and Avoid (DAA) system for unmanned aircraft systems (UAS). The surveillance volume accounts for an airborne sensor capable of detecting non-cooperative aircraft. Independent variables include four candidate DAA Well Clear (DWC) definitions and five surveillance volumes. Open-loop alerting performance metrics are computed from the results of running a reference DAA algorithm on a large number of synthesized encounters. The speed range for the UAS traffic considered is between 40 and 100 kts. Results show that, with a 2.5 nmi sensor range, all four candidate DWCs allow at least an average of 25 seconds warning alert times before a loss of DWC. Cumulative distributions of the intruder's bearing and elevation at the first warning alert suggest that ±10 • and ±140 • , respectively, are sufficient for alerting > 95% of the encounters that lead to losses of DWC.

Detect-and-Avoid Alerting Performance for High-Speed UAS and Non-Cooperative Aircraft

“…Previous work has led to a recommendation that the Low C-SWAP radar have a horizontal field of regard of +/-140 degrees from centerline [16], instead of the Phase 1 value of +/-110, with the goal of allowing 95% of encounters to have their maximum possible alert duration. While the aggregate data presented in Fig.…”

“…The second category consists of four candidate DWC volumes that were being tested for implementation in "Phase 2," and are discussed in detail in Ref. 16. The threshold values of the DWC definition parameters for each of the five DWC's are given in the table below.…”

“…The experiment that underpins this paper leveraged the simulation architecture that was used in a previous study [16]. The data come from a fast-time simulation that used a full day of NAS-wide real VFR traffic data and projected UAS missions and trajectories.…”

“…It is estimated that roughly 15% of VFR traffic is non-cooperative [12], but due to the low number of non-cooperative VFR tracks that could be extracted, some of the lower speed cooperative VFR tracks were used as surrogates for noncooperative traffic in this study. This substitution was made in the previous study [16], and is considered reasonable due to the similarity in the airspeed, turn rate, and accelerations of both cooperative and non-cooperative VFR aircraft when true airspeeds are below 170 knots [18]. The speed distributions of the UAS and VFR traffic are shown below in Fig.…”

A Quantitatively Derived NMAC Analog for Smaller Unmanned Aircraft Systems Based on Unmitigated Collision Risk

“…Bertam and Wei [44], for example, use a definition of a loss of separation of 328 feet (100 meters). Additionally, numerous publications use the NMAC language but fail to define it [45,46,47,48,49,50,51,52].…”