Continuum robots are well suited to medical procedures in deformable anatomical environments, where robotic assistance has the potential to reduce the workload of clinical teams and improve safety. Among the medical procedures, a relevant application is robotic endoscopy, where flexible instruments must navigate confined anatomical passages while adapting to their geometry to the environment. In particular, the present work investigates the robotic actuation and control of a flexible colonoscope.
However, precise control of flexible endoscopes remains challenging because of nonlinear mechanical effects, including backlash, hysteresis, cable slack, variable dead zones, and coupling between actuation channels. These effects complicate the relationship between motor commands and distal tip motion, motivating the development of experimental platforms for modelling and control. This thesis focuses on the development and validation of a robotic platform for investigating these challenges.
The proposed platform employs an Olympus flexible endoscope with four degrees of freedom. The modelling framework aims to describe the relationship between actuation inputs and the endoscope configuration while accounting for uncertainties in the system behaviour.
The proposed methodology also includes a coordinate transformation scheme to account for axial rotation when displaying images acquired by the endoscopic camera. Visual feedback is incorporated into the closed-loop control framework by using the detected lumen centre as a navigation target. Finally, the platform is intended to support the experimental comparison of different navigation strategies, providing a framework for investigating the combined use of kinematic modelling, tip pose measurements, and image-based feedback in robotic endoscopy
