AIDA3 - Rocket: AIDA3 Research AI for Autonomous Aviation - Vertically Integrated Projects

AIDA3 - Rocket: AIDA3 Research AI for Autonomous Aviation

AIDA3 Research: This research project explores and develops machine-learning models to solve real-world problems in fixed-wing uncrewed aerial vehicles (UAVs) and participates in national and international competitions to advance the future of autonomous aviation that are organized both virtually and physically in your unique testing. The project is open to students from all colleges and majors (assuming that they have some basic programming and machine learning skills or have interest in organizing large scale technical competitions).

Advisor:

Sabine Brunswicker

Detailed Description

This VIP research program develops machine-learning models for fixed-wing UAVs and participates in national and international competitions—both virtual and physical—to advance autonomous aviation. Students will organize themselves in smaller teams that fall in the program’s core research thrusts, including (1) perception, navigation, control, as well as (2) digital twin and simulation but also (3) have the opportunity to work on projects outside of these thrusts in the broader area of AI for autonomous aviation (e.g. human autonomy teaming). Teams are encouraged to explore and establish novel methodologies, such as nonlinear model predictive control, reinforcement learning, physics-informed ML, generative AI (LLMs) or embodied AI. Team members also have the unique opportunity to participate in the world’s first international autonomous fixed-wing drone racing competition series and win prizes. Typical activities include software development in Python/C++/ROS 2, training and benchmarking models, building and testing simulation data pipelines, and participating in competitions. Over multiple semesters, teams aim to compete in different national and international competition events and create tools, datasets, and publications for autonomous aviation.