Controlling Soft Robots for Advanced Adaptivity and Versatility in Confined Spaces

Access to confined spaces is highly crucial, whether for repairing hard-to-reach aircraft engine parts, performing minimally invasive surgeries, or inspecting infrastructure after severe weather events.

The Soft-Scout (Controlling Soft Robots for Advanced Adaptivity and Versatility in Confined Spaces) project addresses these challenges by enabling a soft-bodied robot to autonomously navigate and accurately conduct tasks within confined areas, which leads to reduced human effort in these demanding environments and improved quality of inspection and maintenance.

To address the complexity of a soft robotic system, the overarching objective of Soft-Scout is the development and deployment of the control and learning frameworks for a soft robot that provide guarantees for the robot’s precision and reliability, while equipping it with the autonomy and robustness to navigate unknown confined areas.

Maja Trumic will carry out the Soft-Scout project under the supervision of Dr. Cosimo Della Santina at the Delft University of Technology, The Netherlands.

The project is aligned with the United Nations Sustainable Development Goals and the European objectives (European Strategic Agenda and Horizon Europe Mission).

Compared to state-of-the-art solutions, the elasticity of a soft robot’s body allows safe navigation through confined areas, while the proposed framework enables robust and dexterous operation with guaranteed precision and reliability margins.

The Soft-Scout’s vision will be accomplished through three specific objectives: 1) an advanced shape control framework will enable a soft robot to perform various tasks; 2) the twofold learning strategy for autonomous navigation and robustness in confined space will endow a soft robot with the capability to deal with uncertainties in the real-world settings; and 3) the overall control and learning framework will be experimentally validated on a tentacle-shaped soft robot, collecting and integrating user feedback as well.