Investigation of gear design parameters for soft-gear-driven self-propelled colonoscopy robot

Colonoscopy is the most effective method for colorectal cancer screening; however, successful insertion requires a high level of operator skill and may cause patient discomfort or injury. To address these challenges, self-propelled colonoscopy robots that generate propulsion at the distal tip have been actively investigated. In our previous work, we developed a self-propelled colonoscopy robot using a worm-gear mechanism; however, the use of rigid gears posed a potential risk of excessive contact force on the colonic wall.

In this study, we propose a self-propelled colonoscopy robot enabled by soft gears that combine compliant contact with gear-based power transmission. By reinterpreting the crawler-based propulsion concept as a soft spur gear driven by a worm-gear mechanism, the proposed design allows elastic deformation of the gear teeth under load, providing adaptive traction through compliant deformation of the gear teeth. Gear parameters were optimized using gear-design software, and four soft-gear configurations were fabricated to investigate their propulsion characteristics.

The prototypes were fabricated using stereolithography 3D printing and evaluated using a colonoscopy training model. The robot achieved a maximum propulsive speed of 9.6 mm/s and a propulsive force of 1.5 N in straight sections. Gears with larger modules produced higher propulsion performance. In addition, the proposed coaxial double-worm-gear architecture enabled traversal of curved sections of the colon model under manual steering.

These results demonstrate the mechanical feasibility of using soft gears as a propulsion mechanism for self-propelled colonoscopy robots. Although the evaluation was limited to a colonoscopy training model, the proposed design demonstrated stable propulsion in a colonoscopy training model while maintaining compliant contact with the intestinal wall. The simple and compact mechanical structure provides a promising proof of concept for future minimally invasive robotic endoscopy, although further ex vivo and in vivo validation will be required.