Duration: 14 months
This project proposes the development and experimental validation of a mechanically intelligent, tendon-driven actuator–joint architecture for space robotics. The core concept is to shift part of interaction robustness from control algorithms to the actuator itself through intrinsic mechanical compliance, remote tendon routing and physical decoupling of actuator mass from the joint. Embedded elasticity at the transmission level allows the system to absorb geometric misalignment, tolerances and unmodelled disturbances before they propagate into damaging forces.
Beyond compliance, remote tendon routing physically removes motor/gearbox mass from the distal joint, reducing joint-side inertia and contact force amplification which is an advantage that classical SEA architectures do not provide when the actuator remains co-located with the joint.
The architecture combines a BLDC-based drive unit with routed tendons, compliant joint mechanics and embedded proprioceptive sensing, enabling predictable backdrivability and shock tolerance as inherent physical properties. In contrast to classical rigid or series elastic actuators used in terrestrial robotics, the proposed solution is explicitly designed for mechanical intelligence under NACCV conditions, prioritising deterministic behaviour and uncertainty tolerance.
NACCV: non-atmospheric, cold, contaminated and variable-geometry environments typical for lunar surface and on-orbit servicing interaction tasks.
Starting from TRL 3, the activity targets TRL 4 through the design and validation of a tendon-driven joint demonstrator representative of robotic arms for landers, rovers and in-orbit servicing systems, supported by laboratory and screening-level environmental tests at ESA Phi-Lab Poland. The outcome will be a validated actuator–joint concept and reusable architectural baseline for future ISAM missions, lunar surface operations and space infrastructure servicing.
