<p>This paper studies a minimum-time surveillance–evasion differential game between an omnidirectional aerial vehicle (OAV) and a differential-drive robot (DDR). The problem is motivated by autonomous aerial surveillance applications, where UAVs must monitor ground robots under sensing and maneuverability constraints. The OAV (pursuer) maintains a constant altitude and controls a circular detection region corresponding to the footprint of a downward-looking conic sensor, while the DDR (evader) seeks to escape the region in the minimum time. The interaction is formulated as a zero-sum differential game in both realistic and reduced coordinate spaces. Using Pontryagin’s maximum principle and retro-time integration, we derive time-optimal motion strategies and closed-form expressions for the resulting trajectories. The analysis reveals distinct motion regimes separated by singular structures, including a transition surface (TS), where the evader switches controls, and a dispersal surface (DS), where multiple optimal control choices coexist. We characterize the usable part of the terminal set and analyze how speed and geometric ratios affect players’ optimal behavior. Additionally, a feedback-based implementation of the evader’s strategy is developed, enabling state-feedback escape control synthesis. Numerical simulations illustrate the theoretical results, the evolution of players’ trajectories, and compare time-optimal motion strategies against a case where the pursuer employs a suboptimal strategy based on the line-of-sight (LOS).</p>