The Role of Data Science and Machine Learning in Military Aircraft Threat Detection Systems

Abstract

Data science (DS) and machine learning (ML) have become increasingly relevant in the advancement of military aviation threat detection systems (TDS). As enemy threats aimed at destroying aircraft and their aircrew have become more advanced and undetectable, the need for constant improvements to threat detection systems (TDSes) to protect aircraft and aircrew grows. This paper investigates the evolution of threat detection systems from the World War II era to today’s fifth-generation fighters. It examines the implementation of DS and ML in various TDS sensors, including ultraviolet (UV), infrared (IR), radar, and laser detection systems. We also explore the types of threats that TDSes face. These threats include air-to-air and surface-to-air missiles fired from adversarial aircraft and ground-based weapons systems. Lastly, we investigate future applications in TDS with improved technologies, including upgraded sensors for improved differentiation of threats from background clutter and deep reinforcement learning (DRL) for autonomous decision-making in combat scenarios.