Assessment of UAS Detect-and-Avoid Robustness Under Sensor Degradation, Erroneous Inputs, and Interference Scenarios

Abstract

Uncrewed aircraft systems require reliable detect-and-avoid functionality to support operations in airspace shared with crewed aircraft, other uncrewed vehicles, and complex environmental clutter. Contemporary detect-and-avoid systems integrate multiple sensing modalities, onboard navigation, and decision logic to ensure separation standards and mitigate collision risks across a wide range of encounter geometries. However, the operational envelope of these systems is shaped by non-ideal conditions, including gradual sensor degradation, transient faults, corrupt or inconsistent surveillance inputs, and intentional or unintentional interference in sensing and communications channels. Understanding how such conditions propagate through tracking, conflict detection, and maneuver selection is important for evaluating safety margins and for informing system design choices. This paper develops a structured assessment of detect-and-avoid robustness under these conditions, focusing on the interaction between sensing imperfections, estimator performance, conflict prediction uncertainty, and guidance decisions. The analysis considers heterogeneous sensor architectures, probabilistic models of degradation and faults, and interference mechanisms that perturb either measurement streams or the logical integrity of detect-and-avoid functions. Emphasis is placed on characterizing conditions under which detect-and-avoid performance degrades gradually, conditions under which it collapses abruptly, and the sensitivity of critical safety metrics to modeling assumptions. The resulting formulations and illustrative evaluations provide a basis for comparing detect-and-avoid configurations under stress, identifying parameter regimes of concern, and informing verification activities that incorporate adverse but plausible sensing and interference scenarios.