Oklahoma State Researcher Receives $899K NSF Grant to Advance AI-Enabled Drone Autonomy
Researchers at Oklahoma State University (OSU) are developing new technologies that combine artificial intelligence, advanced control systems, and human supervision to make drones more autonomous and effective in complex inspection tasks. The project is supported by a $899,000 National Science Foundation (NSF) grant awarded to Yafeng Wang, assistant professor in OSU's School of Materials, Mechatronics and Manufacturing Engineering.
The project, titled CPS-FR: A Cyber–Physical–LLM–Human System (CPLHS) Framework for Vision-Based Drone Applications, aims to transform drones from remotely controlled tools into intelligent robotic partners capable of interpreting mission requirements, collecting useful visual information, and adapting to changing conditions.
The proposed framework integrates two AI components. Supervising AI uses engineering knowledge and mission requirements to guide operations, while Field AI manages visual data collection and responds to challenges encountered during flight. Human engineers remain involved by defining mission objectives, monitoring performance, and intervening when necessary.
A key research objective is improving the quality of images collected during drone inspections. Rather than following fixed flight paths, drones could recognize problems such as glare or motion blur and automatically adjust their position, speed, or camera angle to obtain better results.
The team is also developing models that account for interactions between drone movement and camera gimbal positioning, along with adaptive control techniques that balance competing mission objectives while maintaining stable flight.