ABSTRACT
The emergence of the Internet of Things (IoT) in recent years has revolutionized the whole technological sphere through diverse applications in different fields from manufacturing to health care. A large scope is created in every field, especially in health care, which reserves the most market share among all such smart applications. The concepts of tele-health, e-health, m-health etc., along with their association with cyber-physical systems (CPSs), are guiding today's global healthcare industry. As a significant part of it, we are also witnessing innovative attempts for the betterment of the quality-of-life of human beings, especially the elderly population, which is following an unprecedented rise in today's world. In this regard, we propose the development of an end-to-end CPS named I-ACTIVE, which ensures ubiquitous recognition of common human daily living activities in real time and with less cost in comparison with smartphone-based human activity recognition (HAR) systems, without compromising the recognition efficiency. The I-ACTIVE prototype uses an accelerometer as the primary sensing component; to enable wireless connectivity to the Internet it relies on ESP8266 Wi-Fi module. It exploits Google cloud for necessary computation and recognizes human activity in real time. The I-ACTIVE system yields a 97.71% accuracy for complex daily life activities, which is further extended to 98.5% in the cases of simple and standard basic activities.
