Smart Forests: fire detection service

Abstract

Smart Forest is a concept derived from the Internet of Things (IoT) and defines sections of a forest where remote sensing is applied to collect data about environmental conditions. One of the main objectives of Smart Forests is to detect wild fire at early stages. However, the required technology for such monitoring usually demands a complex and expensive sensor and network infrastructure and requires central processing capabilities for analyzing data from several thousands of sensors. The goal of this work is to propose a solution focused on Edge Computing, using the concept of Mobile Hubs (M-Hubs). The developed Fire Detection IoT prototype application is based on the ContextNet middleware, and uses Event Processing Agents (EPAs), running on smart phones carried by forest guards. Scalability tests up to 10,000 remote sensors connected per MHub were performed.