An Automatic and Accurate Localization System for Firefighters

Abstract

Firefighters' safety is a critical problem and a major issue is the lack of reliable indoor firefighter localization. State of the art approaches have failed to provide an automatic, accurate and reliable solution to localize firefighters in harsh environments. This paper presents a novel system to achieve this goal, by combining pedestrian dead reckoning with a recently emerging breadcrumb system. Our solution includes a new collaborative localization algorithm that contains a novel marginalization scheme and can improve the location accuracy of firefighters. We fully implement the algorithm in a complete system and conduct experiments in both an office building and in a simulated firefighting scene that involved a real fire and professional firefighters. Evaluation results from a 400 meter-long trace demonstrate that our approach significantly reduces the average and maximum firefighter location error to 1.4% and 2.7% of the total distance, respectively.