Context-aware energy optimization for perpetual IoT-based safe communities

• Published in: Sustainable computing informatics and systems 2019-06, Vol.22, p.96-106
• Main Authors: Alhassoun, Nailah Saleh, Sarwar Uddin, Md Yusuf, Venkatasubramanian, Nalini
• Format: Article
• Language: English
• Subjects: Activities of daily living; Energy optimization; Internet of things; Sensors’ activation scheduling algorithm; Smart communities; Wireless sensors network
• ISSN: 2210-5379
• DOI: 10.1016/j.suscom.2019.01.016

•Perpetual awareness systems are sensing systems characterized by continuous monitoring and ubiquitous sensing; they are essential to many safety and mission-critical applications, e.g. assisted living, healthcare and public safety.•The SAFER system for assisted living, uniquely leverages context information (e.g. activities of daily living ‘ADLs’, location) for energy-optimized sensor activations to detect anomalous events, such as falls.•Experimental studies with real world datasets indicated that the proposed technique was able to achieve more than 80% reduction in energy consumption, in a heterogeneous IoT platform for fall detection, thereby doubling the system-lifetime without loss of sensing accuracy. The IoT revolution has provided a promising opportunity to build powerful perpetual awareness systems. Perpetual awareness systems are sensing systems characterized by continuous monitoring and ubiquitous sensing; they are essential to many safety and mission-critical applications, e.g. assisted living, healthcare and public safety. In this paper, we present SAFER, a perpetual heterogeneous IoT system; deployed in homes to detect critical events (injury, hazardous-environment) that must trigger immediate action and response. A key challenge here is the energy consumption associated with perpetual operations. We propose a novel energy-aware perpetual home IoT system where battery-operated and wall-powered IoT devices co-execute to ensure safety of occupants. We use a semantic approach that utilizes context of extracted activities of daily living (ADLs) from device data to drive energy optimized sensor activations. To validate our approach, we developed an elderly fall detection system using multi-personal and in-situ sensing devices derived from real world deployments in SCALE project which has been deployed in Montgomery county, MD. Using initial measurements to drive larger simulations, we show that proposed Cost-Function-Gradient algorithm can achieve greater than 4X reductions in energy dissipation and doubling system of battery powered devices lifetime without loss of sensing accuracy.