Abstract
In this work, we aim to address the reliability issues in community-wide large-scale Internet-of-Things (IoT) deployments. Traditional IoT systems depend heavily on infrastructure thus are difficult to scale up and operate reliably to cover wide areas like communities and cities, leaving them vulnerable to dynamics in long-term operation. Given the increasing number of mobile devices with sensing capabilities in communities, we argue that the reliability of community IoT systems can be improved by leveraging mobility. We highlight essential research challenges in building integrated community IoT ecosystems with mobile nodes, and propose several possible approaches to improve sensing coverage and information quality. Based on our experience in building resilient IoT middleware systems, we designed SCALECycle, a mobile sensing and data collection platform applicable in bike-friendly communities. We illustrate our initial experiences through measurement studies on three real-world testbeds (in the US and Taiwan) and show the value of intelligent scheduling policies through extensive simulations.