Vision-Guided Aerial Manipulator

Autonomous grasping and placement using quadrotor motion planning and vision

This project focused on enabling a quadrotor to perform grasp-and-place tasks with visual feedback. The system combined motion planning, control, and perception so the aerial platform could execute smooth, stable manipulation actions.

What I worked on

  • Implemented minimum-snap trajectory optimization for smooth quadrotor motion profiles.
  • Integrated motion controllers with VLM outputs to support LiDAR-free autonomous manipulation.
  • Stabilized grasping and placing behaviors using the manipulator system.

Key takeaway

The project demonstrated how visual guidance and optimized motion generation can be combined to enable semi-autonomous aerial manipulation in practical scenarios.