Smart Glove Calibration and Data Retrieval

Abstract

Work related health issues prevent people from long term service and cause huge financial loss for the companies and individuals. To avoid these health issues, ergonomic studies are being conducted and preventive measures are being taken. Musculoskeletal injuries due to hand movements can be prevented by using a smart glove which can indicate the risk conditions. The smart glove used in this thesis is a glove knitted with Bekinox 50/2 conductive yarns at different parts of it to identify the risk condition. Even though a textile sensor has many advantages, its calibration is a major problem. This is mainly due the deviation caused by textile orientation, inhomogeneity, differences in hand structure of the wearer etc. Therefore the thesis mainly studies the textile-sensor characteristics in order to examine if it can be accurately calibrated with minimum recalibration requirements for the smart glove application. The results show that the current glove, with the force sensor knitted using Bekinox 50/2 can only be used for low force measurement and is not very well suited for smart glove application. It is found that the sensor’s repeatability can be improved by providing a stable structure. The wrist-angle sensor in the current glove should be incorporated with more sensors to identify the sideways movement to use it for smart-glove application. An efficient method of communicating sensor values through the Internet was implemented in the thesis using the MQTT protocol. Also, Python scripts for plotting real-time graph and storing data were implemented successfully.