A Mobile Application for Wooden House Fire Risk Notifications Based on Edge Computing

We investigate how recent advances in computational modelling of fire risk for single-unit wooden houses can be used as a foundation for a mobile application capable of notifying users of in-home fire risk. Our approach relies on publicly available cloud services providing historic weather data and weather forecast data to compute fire risks for geographical locations on a mobile (edge) device. We undertake an iterative usability study focusing on the user interface to determine how fire risk must be conveyed in a form suitable for end-users. Furthermore, we evaluate CPU and memory consumption to assess the computational feasibility of implementing the fire risk computations directly on mobile devices. Our contribution is a validated design of a user interface for a mobile fire risk application, and a demonstration of fire risk computations being feasible on mobile devices.