REAM: A Framework for Resource Efficient Adaptive Monitoring of Community Spaces

Abstract

Nowadays, many Internet-of-Things (IoT) devices with rich sensors and actuators are being deployed to monitor community spaces. The data generated by these devices are analyzed and turned into actionable information by analytics operators. In this article, we present a Resource Efficient Adaptive Monitoring (REAM) framework at the edge that adaptively selects workflows of devices and analytics to maintain an adequate quality of information for the applications at hand while judiciously consuming the limited resources available on edge servers. Since community spaces are complex and in a state of continuous flux, developing a one-size-fits-all model that works for all spaces is infeasible. The REAM framework utilizes reinforcement learning agents that learn by interacting with each community space and make decisions based on the state of the environment in each space and other contextual information. We demonstrate the resource-efficient monitoring capabilities of REAM on two real-world testbeds in Orange County, USA and NTHU, Taiwan, where we show that community spaces using REAM can achieve >90% monitoring accuracy while incurring ∼50% less resource consumption costs compared to existing static monitoring approaches. We also show REAM’s awareness of network link quality in its decision-making, resulting in a 42% improvement in accuracy over network agnostic approaches.