Wireless sensor networks for activity monitoring in safety critical applications
R. Rednic, E.I. Gaura, J. Brusey
Coventry University, UK
Keywords: posture, bsn, accelerometers
Abstract:The ability to monitor posture is essential to many application areas, including virtual reality, health, and sports applications. The work here focuses on the use of postural monitoring monitoring in safety critical missions such as explosive ordnance disposal (EOD) missions. The operatives undertaking these missions are commonly placed under a high level of physical and psychological strain due to the weight of the protective armoured suit and the importance of their work. Towards enhancing the operatives safety in such missions, remote monitoring of posture may allow a better understanding of the operative’s status, particularly when combined with additional health information. Previously, a Body Sensor Network-based (BSN) posture monitoring system consisting of nine accelerometers placed strategically on a human subject’s body was designed and implemented by the authors here. The system was able to recognise six specific postures (sitting, kneeling, crawling, and three variations of laying on the ground) with up to 100% accuracy. However, due to limitations of the data processing method used, the system was unable to consistently distinguish between a subject standing, walking and running - an essential differentiation required by the application at hand. In order to counteract this limitation, a new prototype utilising additional sensors and an augmented data processing method has been implemented and evaluated and is reported here.