AI-based Smart System

home > Research > AI-based Smart System > b. Vision-based Automated Aerial Docking System
b. Vision-based Automated Aerial Docking System

Principal Investigator: Jae-Hung Han
Participating Graduate Research Assistants: Jaeyong (Andrew) Choi, Jeonghwan Park
Related Projects: KEIT (한국산업기술평가관리원 / 항공우주부품개발사업)

Related Recent Publication:

  • [1] Choi, A.J., Park, J., and Han, J.-H., “Automated Aerial Docking System Using Onboard Vision-Based Deep Learning,” AIAA Journal of Aerospace Information Systems, Vol. 19, No. 6, pp. 421-436, Jun. 2022.
  • [2] Choi, A. J., Park, J., and Han, J.-H., “Automated Aerial Docking System using Vision-Based Deep Learning,” SciTech 2022, San Diego, CA & Virtual, Jan. 3-7, 2022.
  • [3]Choi, A. J., Park, J.-H., and Han, J.-H., “Development of bi-stable based aerial integration and separation mechanism.” The 2021 Asia-Pacific International Symposium on Aerospace Technology, Jeju, Korea, Nov. 15-17, 2021.

1.Goals

Develop automated aerial docking system using onboard vision-based deep learning
  • Bi-stable type of aerial docking/releasing mechanism.
  • Implement vision-based deep learning target detection/tracking system.
  • Performance validation of the developed automated aerial docking system for the reliability and robustness.

2. Approaches

Bi-stable type of aerial docking/releasing mechanism
  • Stable in both locked/docked and released states.

Implement vision-based deep learning target detection/tracking system
  • System that can be operated by onboard ML computing module.

Performance validation of the developed automated aerial docking system for the reliability and robustness.
  • Ground test using robot arms & indoor flight test using quadcopter drone