Efficient Real-time Smoke Filtration with 3D LiDAR for Search and Rescue with Autonomous Heterogeneous Robotic Systems

TL;DR

The paper addresses the challenge of aerosol-induced noise in LiDAR data for search-and-rescue missions in harsh, GNSS-denied environments. It introduces a modular, platform-agnostic filtration pipeline that combines radius-based filtering, intensity-based outlier removal using a Weibull-derived threshold, dynamic cluster-based outlier rejection, Savitzky–Golay smoothing, and 2D radius-based pruning, with close- and long-range segmentation to balance accuracy and latency. Implemented in Python on CPU, the framework adapts to platform velocity and environmental density, achieving real-time performance (10–20 Hz) and demonstrably improving obstacle detection and map quality in smoke-filled field tests with heterogeneous robotic teams. The work supports safer autonomous navigation in SAR scenarios and suggests avenues for further optimization, including alternative clustering methods and a C++ implementation for additional performance gains.

Abstract

Search and Rescue (SAR) missions in harsh and unstructured Sub-Terranean (Sub-T) environments in the presence of aerosol particles have recently become the main focus in the field of robotics. Aerosol particles such as smoke and dust directly affect the performance of any mobile robotic platform due to their reliance on their onboard perception systems for autonomous navigation and localization in Global Navigation Satellite System (GNSS)-denied environments. Although obstacle avoidance and object detection algorithms are robust to the presence of noise to some degree, their performance directly relies on the quality of captured data by onboard sensors such as Light Detection And Ranging (LiDAR) and camera. Thus, this paper proposes a novel modular agnostic filtration pipeline based on intensity and spatial information such as local point density for removal of detected smoke particles from Point Cloud (PCL) prior to its utilization for collision detection. Furthermore, the efficacy of the proposed framework in the presence of smoke during multiple frontier exploration missions is investigated while the experimental results are presented to facilitate comparison with other methodologies and their computational impact. This provides valuable insight to the research community for better utilization of filtration schemes based on available computation resources while considering the safe autonomous navigation of mobile robots.

Figures (8)

• Figure 1: Behavior of in the presence of dust, where the propulsion system of a drone entrains dust particles.
• Figure 2: Test 1. Row 1: Generated map. Left, original , middle, filtered and isolated dust, right. Row 2: Weibull distributions for the s.
• Figure 3: Overview of the filtration framework.
• Figure 4: Challenging environment for sensors.
• Figure 5: Test 2: map generation with and without our filtration framework.