Aerial Manipulation with Contact-Aware Onboard Perception and Hybrid Control
Yuanzhu Zhan, Yufei Jiang, Muqing Cao, Junyi Geng
AI summary
Problem
Most aerial manipulation demonstrations rely on external motion capture and position-only control, limiting real-world deployability. Standard onboard odometry lacks the accuracy needed for simultaneous motion and force regulation during contact, while perception–control coupling destabilizes interactions.
Approach
The method augments visual–inertial odometry with contact-consistency factors that activate only during interaction to reduce drift, pairs it with image-based visual servoing to mitigate perception–control coupling, and employs a hybrid force–motion controller to regulate contact wrenches and lateral motion.
Key results
- 66.01% improvement in velocity estimation at contact via contact-aware VIO
- Stable contact wrench regulation and precise lateral tracking using hybrid force–motion control
- Fully onboard perception-to-wrench loop closed without external motion capture or GPS
- Validated in simulation and real-world experiments with reliable target approach and force holding
Why it matters
Enables deployable, contact-rich aerial manipulation in GPS-denied or infrastructure-rich environments, advancing maintenance, agriculture, and inspection applications.
Abstract
Aerial manipulation (AM) promises to move Un- manned Aerial Vehicles (UAVs) beyond passive inspection to contact-rich tasks such as grasping, assembly, and in-situ maintenance. Most prior AM demonstrations rely on exter- nal motion capture (MoCap) and emphasize position control for coarse interactions, limiting deployability. We present a fully onboard perception–control pipeline for contact-rich AM that achieves accurate motion tracking and regulated con- tact wrenches without MoCap. The main components are (1) an augmented visual–inertial odometry (VIO) estimator with contact-consistency factors that activate only during interac- tion, tightening uncertainty around the contact frame and reducing drift, and (2) image-based visual servoing (IBVS) to mitigate perception–control coupling, together with a hy- brid force–motion controller that regulates contact wrenches and lateral motion for stable contact. Experiments show that our approach closes the perception-to-wrench loop using only onboard sensing, yielding an velocity estimation improvement of 66.01% at contact, reliable target approach, and stable force holding—pointing toward deployable, in-the-wild aerial manipulation.