Abstract
Globally, there is a growing number of individuals living with amputations due to the increasing incidences of diabetes and vascular diseases in an aging population. A prosthesis seeks to restore some of the functional elements of the missing limb, helping the user improve their quality of life. One of the important components of a prosthesis is the prosthetic socket, where its fit is a key determinant of the prosthesis’ success or failure. For decades, evaluation of socket fit has been done through clinical assessment methods, which heavily relied on tacit knowledge and is subjective in nature. The work presented in this thesis details the design and use of a soft sensor-based socket interface which is able to evaluate parameters of socket fit under dynamic conditions. The data obtained from the sensors could help provide quantification of important parameters such as pressure and volume changes within the socket. This would allow both prosthesis users and prosthetists to understand and make better informed decisions regarding socket fit. Beginning with a review of related work in the area, insights and challenges were drawn from sensing technologies that had the potential to be used within the prosthetic socket. These technologies were compared for relevance and suitability. A design discovery exercise was then carried out to obtain a clearer understanding of the needs and expectations of the sensor system’s potential users. Insights gained from this stage aided in generating the specifications and technical requirements of the sensing socket interface. The prototype design was then conceptualized and developed through an iterative process, where the sensors were refined and tested to evaluate its workability. Lastly, a pilot study was done to test the sensing socket interface in real-time, within prosthetic sockets of three prosthesis users. The results of the test were analysed and feedback from users were collected on sensor performance and the potential benefits of having such a system. Findings were discussed and insights were drawn to highlight areas of improvement in future work. In summary, it was found that use of the soft sensor-based socket interface was positive amongst surveyed users and results demonstrate its validity in sensing and improving quantification of socket fit parameters. Future work could look into improving sensor resolution and also potential applications with actuators and control loops to develop self-adjusting sockets.