XAI-BattleOps: Explainable AI-based Energy Consumption Analytical Framework for UAV-assisted Battlefield Operations
Khushi Shah, Henil Shalin Panchal, Lakshin Pathak, Nilesh Kumar Jadav, Sudeep Tanwar, Nagendar Yamsani · 2024
In the rapidly evolving landscape of unmanned aerial vehicles (UAVs), it is crucial to optimize energy consumption for extending its operational capabilities and improving efficiency. In this study, we look at how explainable artificial intelligence can be used to predict the energy consumption of UAVs in military operations. Conventional artificial intelligence (AI) black-box models are powerful but provide no visibility into their judgment calls, creating mistrust and restricting regulation compliance. Explainable artificial intelligence (XAI) techniques were utilized in this paper to develop interpretable models for predicting UAV energy consumption based on various operational parameters. These models provide valuable insights into factors influencing energy usage, thereby enabling stakeholders to make informed decisions, optimize flight plans, and adopt energy-saving measures. Using an extensive dataset, we assess the robustness and accuracy of XAI methodologies. We found out from our analysis that XAI maintains predictive performance and improves model interpretability, thus increasing adoption trust in AI-enabled energy management solutions for UAVs in military applications.