Soft Variable Stiffness Tactile Sensors for Clinical Examinations

Abstract

Emerging tactile robotic technology gives robots the ability to sense exclusively tactual properties of objects in their environment, such as texture and softness, extending the reach of robotics into applications that require high dexterity. Using sensors that are made of soft materials means that robots can interact delicately and safely with their environment. Soft, tactile sensors may therefore be designed for applications that are traditionally executed by people. For example, palpation of breast tissue by healthcare professionals is widely and frequently performed, yet reports of its effectiveness, and of competence in performing an examination, vary greatly. This thesis presents the development of a soft, variable stiffness tactile sensor and manipulator platform for the purpose of performing clinical breast examinations. The development is guided and corroborated by involvement from the intended end-users.

First, a pneumatic-elastomeric variable stiffness mechanism is introduced and characterised through modelling and experimentation. Shape change of the elastomeric membrane due to pneumatic actuation is localised to each taxel, minimising global membrane deformation. Tactile cues acting on the variable stiffness membrane are hydraulically translocated to a transduction mechanism at another location. This translocation greatly frees up the parameters for the transduction, hence a significant contribution of this thesis is an exploration of multiple transduction mechanisms for this hydraulic signal. Then, a variable stiffness soft sensor is demonstrated to be able to detect small surface details, classify subtly different stiffness objects, and estimate the size and location of deeply embedded nodules. The basis of the sensing mechanism is shown to be scalable in terms of density, taxel size, and taxel number. Following this demonstration, a robotic manipulator is developed, and demonstrated in simulation and experiments to be capable of holding multiple sensors to cover a breast surface and apply forces consistent with a comfortable medical examination. Finally, the thesis is contextualised in terms of the intended end-users’ acceptance of the palpation platform service.

The findings of this thesis have potential applications in healthcare, as well as other applications where delicate manipulation is required, such as the nuclear industry, manufacturing, assisted living, surgery, and entertainment.

Date of Award	19 Mar 2024
Original language	English
Awarding Institution	University of Bristol
Supervisor	Antonia Tzemanaki (Supervisor) & Andrew T Conn (Supervisor)