Dinorego Mphogo

Dinorego Mphogo is a PhD student whose research applies machine learning to aviation safety systems. His current work focuses on replicating and stress-testing the decision logic of collision avoidance systems using simulated aircraft encounter data, combining air traffic control text data with supervised classification and explainability techniques to understand how data-driven models can emulate deterministic, safety-critical aviation logic.

His research explores both the fidelity of large language models in understanding and reproducing collision-avoidance advisories and their robustness to missing or degraded input features — work with direct implications for the design of next-generation Detect-and-Avoid systems for unmanned aircraft.