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 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 potential risk of their work. Remote monitoring of posture may allow a better understanding of the operative’s status. When combined with additional health information, posture can enhance the accuracy of operative’s global state estimation. Previously, a Body Sensor Network-based (BSN) posture monitoring system consisting of nine accelerometers 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 high accuracy. However, the system was unable to consistently distinguish between a subject standing, walking or running. 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.