ENABLING THE TEST AND EVALUATION OF AUTONOMOUS FUNCTIONS THROUGH STANDARDIZATION OF UAS INTEGRATION

Autonomous capabilities for uncrewed aerial systems (UAS) have advanced rapidly in recent years; however, the transition of autonomous functionality from laboratory environments toreal-world flight testing remains limited by regulatory constraints, hardware compatibility challenges, and the absence of standardized interfaces between autonomy software and flight control systems. These limitations are particularly significant for research programs operating under Federal Aviation Administration (FAA) Part 107 regulations and 2018 National Defense Authorization Act (NDAA) hardware sourcing restrictions. This thesis presents a standardized autonomy-to-flight-controller interface framework designed to support the integration, testing, and evaluation of advanced autonomous behaviors on research UAS platforms. The framework, referred to as ”Jacob’s Ladder,” establishes a modular architecture that separates high-level autonomy functions from low-level flight control while providing a unified communication protocol between the two layers. The system combines a ROS 2-based autonomy stack, PX4 flight control firmware, a Micro XRCE-DDS communication bridge, and a containerized development environment to enable seamless deployment across simulation and hardware platforms. A complete development and validation pipeline was implemented using Gazebo softwarein-the-loop (SITL) simulation, hardware-in-the-loop (HITL) bench testing, and real-world flight experiments. The architecture was deployed on a companion-computer configuration consisting of an NVIDIA Jetson Orin Nano and a Pixhawk Cube Orange Plus flight controller, selected for their compatibility with open-source flight software and NDAA-compliant research platforms. Autonomous mission behaviors were first validated in simulation and bench testing, and were subsequently demonstrated in flight through basic visual navigation tasks in which the aircraft detected and navigated toward a target using onboard perception and autonomy software. The results confirm that the proposed framework provides a robust and portable method for integrating autonomy algorithms with research aircraft while reducing system-integration overhead and improving reproducibility across institutions. By standardizing the communication pathway between autonomy software and flight controllers, the framework establishes a scalable foundation for collaborative autonomy research and enables a structured transition of autonomy algorithms from simulation to real-world flight testing. Future work will focus on expanding the autonomy capabilities supported by the framework and applying the architecture to more complex missions such as shipboard landing and operations in Global Positioning System (GPS)- and Radio Frequency (RF)-denied environments.

Department dc:contributor.department

Aerospace Engineering

Year dc:date.issued

2026

Author dc:creator

• Safeer, Jacob

Advisor dc:contributor.advisor

• Costello, Donald

Language dc:language.iso

en

Identifier

https://doi.org/10.13016/lcmm-hnop

OAI identifier oai:identifier

oai:drum.lib.umd.edu:1903/35517