Closing the Loop: Autonomous UAV Mission Design and Execution with AERPAW

Anirudh Kamath

Graduate Researcher, University of Utah

Anirudh Kamath, a graduate researcher at the University of Utah and team lead of the InFlux team, is advancing how autonomous UAV missions are designed, executed, and optimized in dynamic wireless environments. A central challenge in UAV-based data collection missions is the tight integration of communication planning, UAV control, and data collection strategies into a unified and reliable workflow. Traditional approaches often treat these components separately, limiting adaptability and reducing mission efficiency. To address this, Kamath and his team developed an iterative decision-making approach that continuously determines the optimal destination for the UAV based on the spatial distribution of data volumes across ground stations. Their method evaluates a utility function that accounts for remaining data, signal quality, UAV dynamics, and mission constraints—allowing the UAV to adapt its trajectory in real time and make informed decisions about when to move, where to go next, and how to maximize data collection efficiency.

This approach proved highly effective in practice: Kamath’s team won 1st place in the AERPAW Autonomous Aerial Data Mule (AADM) Challenge, outperforming 15 competing teams in a multi-phase competition that required UAVs to autonomously collect data from multiple base stations under strict time and energy constraints. The winning solution incorporated not only adaptive waypoint selection but also considerations such as UAV speed, geofence restrictions, signal-to-noise ratios, and optimal UAV orientation, all derived and refined through experimentation in AERPAW’s digital twin and validated with real-world insights.

AERPAW was central to enabling this success. The team leveraged the platform’s digital twin to plan flight trajectories, configure UAV behavior programmatically, and simulate realistic communication conditions, allowing them to iteratively refine their strategy before deployment. As part of the broader AADM challenge ecosystem, which combines digital twin experimentation with outdoor testbed validation, the team’s work demonstrates how integrated, iterative design can lead to robust and high-performing autonomous UAV systems .

Through this work, Kamath and his team illustrate the power of combining adaptive algorithms with high-fidelity experimentation infrastructure, showing that intelligent, data-driven UAVs can significantly improve mission efficiency and reliability. Their contributions are also reflected in the broader AERPAW dataset and research effort documented in “UAV-Based Wireless Multi-modal Measurements from AERPAW Autonomous Data Mule (AADM) Challenge in Digital Twin and Real-World Environments,” submitted to IEEE Data Descriptions in February 2026.