Understanding Fire Through Thermal Radiation Fields for Mobile Robots

TL;DR

This work presents a novel approach for mobile robots to understand fire through building real-time thermal radiation fields by using the Stefan-Boltzmann law to approximate the thermal radiation in empty spaces and shows that this representation can be used for robot navigation.

Abstract

Safely moving through environments affected by fire is a critical capability for autonomous mobile robots deployed in disaster response. In this work, we present a novel approach for mobile robots to understand fire through building real-time thermal radiation fields. We register depth and thermal images to obtain a 3D point cloud annotated with temperature values. From these data, we identify fires and use the Stefan-Boltzmann law to approximate the thermal radiation in empty spaces. This enables the construction of a continuous thermal radiation field over the environment. We show that this representation can be used for robot navigation, where we embed thermal constraints into the cost map to compute collision-free and thermally safe paths. We validate our approach on a Boston Dynamics Spot robot in controlled experimental settings. Our experiments demonstrate the robot's ability to avoid hazardous regions while still reaching navigation goals. Our approach paves the way toward mobile robots that can be autonomously deployed in fire-affected environments, with potential applications in search-and-rescue, firefighting, and hazardous material response.

Figures (10)

• Figure 1: Experimental setup with the Boston Dynamics Spot robot approaching a controlled fire generated by a propane-based fire training device. The robot is equipped with depth and thermal sensors to perceive the environment and construct a thermal radiation field for fire-aware navigation.
• Figure 2: Overview of our framework: we use stereo depth and thermal sensor data to capture a fire (a) and to compute a thermally annotated point cloud (b). We use the points with the highest temperatures and project them into a 2D grid (c) and use the Stefan-Boltzmann law to estimate the heat decay of the fire (d). Finally, we integrate the computed thermal occupancy map with a common spatial occupancy map for fire-aware robot navigation (e).
• Figure 3: To register the right (a) and left (b) RGB frames of the Spot robot and of the thermal (c) camera, we use the Calibmar toolkit seegraeber2024. We use a heated ArUco board to obtain a thermal frame with significant thermal differences between the white and black areas.
• Figure 4: Registration result: we show the registered RGB frames of the Spot with a thermal overlay (a) as well as a zoomed in view to show more details (b). The RGB frames have been converted to a grayscale image to show the color coded thermal overlay.
• Figure 5: Experimental setup showing the Boston Dynamics Spot robot and the fire source turned off (a) and turned on (b). The thermal images taken from the perspective of the robot show the corresponding temperatures obtained with the thermal camera (c), (d).