Falls in elderly is one of the most serious causes of severe injury. Lack in immediate medical help makes these injuries life threatening. An automatic fall detection system, presented in this research, would help reduce the arrival time of medical attention, reduce mortality rate and promote independent living. Therefore, the algorithm finds its application in the medical field, specifically in nursing homes. The system designed and presented in this research is not only capable of detecting human falls but also distinguishing them from routine fall-like activities. Falls are detected with the help of a small wearable embedded device, i.e. Texas Instruments' eZ430 Chronos watch which is wireless development kit. The watch operates at an RF frequency of 915MHz to communicate with each other in a wireless network. The wearable wrist watch is programmable and has an in-built accelerometer sensor and microcontroller circuitry. The accelerometer sensor is motion sensitive and measures the acceleration due to gravity. Whenever a fall is detected the watch sends a signal to the neighboring watch, which is always in the monitoring mode. Signal transmission and reception between these devices is via wireless communication, where every node is a sensor forwarding the signal to the next node. A wireless mesh network helps in quick transmission of signals thereby alerting the authorities. In order to differentiate between body fall and Activities of Daily Life, various body motions and gestures have been studied and presented. The features of a real fall and that of normal human motions are extracted and analyzed from the data obtained by volunteers who participated in the research. Evaluation of results led to setting forth threshold values for parameters like acceleration, change in co-ordinate axes and angle of orientation. Over-passing the threshold raises a fall alarm to bring to the attention of the hospital authority.
Library of Congress Subject Headings
Falls (Accidents) in old age--Remote sensing; Wireless sensor networks; Wearable computers--Programming
Department, Program, or Center
Computer Engineering (KGCOE)
Rakhecha, Sanjana, "Reliable and secure body fall detection algorithm in a wireless mesh network" (2013). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus