Integrated tip tracking and whole-body collision avoidance for multi-segment continuum robots based on real-time shape reconstruction

Continuum robots have found widespread applications in clinical surgery, benefiting from their high flexibility and compliance. However, their inherent compact design and passive compliance present significant challenges for precise shape sensing and control, which limit their safe and reliable operation in constrained environments. This paper proposes a collaborative collision avoidance and tracking framework, aiming to precisely follow the pre-planned trajectory without the manipulator body intruding into unsafe region. First, an augmented shape description model is constructed, upon which a novel online shape reconstruction strategy is developed. In this strategy, only the end-effector’s position and direction are required as feedback, thus enabling position estimation of arbitrary points along the manipulator with few sensors. Then, a collision-free robust adaptive control scheme is proposed, which can effectively suppress the effects of uncertainties and external disturbances through real-time Jacobian matrix updates and compensation. Furthermore, a virtual force-based sub-controller is designed within the framework for collision avoidance. By utilizing the Lyapunov-based technique, the stability of the closed-loop system is proven. Finally, a series of validation and comparative experiments were conducted on the test platform. The results demonstrate the feasibility and effectiveness in achieving accurate path tracking while actively avoiding the unsafe region.