Experimental open-source framework for underwater pick-and-place studies with lightweight UVMS – an extensive quantitative analysis

The rise of lightweight, low-cost underwater vehicle-manipulator systems (UVMS) has made autonomous underwater manipulation increasingly accessible. Yet, most current research remains limited to isolated tasks, such as trajectory tracking or compensation of unknown payloads. Detailed experimental analyses that go beyond a proof-of-concept are particularly rare.We present a comprehensive open-source software framework for fully automated pick-and-place studies. We build upon our previous work on a task-priority control framework and extend it to enable fully autonomous manipulation. This includes a high-level decision-making process to coordinate the pick-and-place sequence and a grasp detection method to verify the successful pick-up of the object. We demonstrate this framework on the widely-used platform of a BlueROV2 and an Alpha 5 manipulator.Extensive quantitative experimental studies (100+ trials) show the picking and placing to be highly accurate, with mean position errors of <5 mm and <10 mm, respectively. We additionally validate our grasp detection approach and analyze trajectory tracking sensitivity to varying payloads and speeds. These results provide a baseline of what accuracy is currently achievable with state-of-the-art lightweight hardware under ideal research conditions. The code is available at https://github.com/HippoCampusRobotics/uvms.

Subjects

Accuracy

Sensitivity

Codes

Trajectory tracking

Statistical analysis

Decision making

Hardware

Open source software

Intelligent robots

Payloads

DDC Class

551: Geology, Hydrology Meteorology

620.1: Engineering Mechanics and Materials Science

Options

Experimental open-source framework for underwater pick-and-place studies with lightweight UVMS – an extensive quantitative analysis