SPACE: 3D Spatial Co-operation and Exploration Framework for Robust Mapping and Coverage with Multi-Robot Systems

TL;DR

A new semi-distributed framework for spatial cooperation in indoor environments that enables enhanced coverage and 3D mapping and introduces a novel spatial frontier detection system and map merger, integrated with an adaptive frontier assigner for optimal coverage balancing the exploration and reconstruction objectives.

Abstract

In indoor environments, multi-robot visual (RGB-D) mapping and exploration hold immense potential for application in domains such as domestic service and logistics, where deploying multiple robots in the same environment can significantly enhance efficiency. However, there are two primary challenges: (1) the "ghosting trail" effect, which occurs due to overlapping views of robots impacting the accuracy and quality of point cloud reconstruction, and (2) the oversight of visual reconstructions in selecting the most effective frontiers for exploration. Given these challenges are interrelated, we address them together by proposing a new semi-distributed framework (SPACE) for spatial cooperation in indoor environments that enables enhanced coverage and 3D mapping. SPACE leverages geometric techniques, including "mutual awareness" and a "dynamic robot filter," to overcome spatial mapping constraints. Additionally, we introduce a novel spatial frontier detection system and map merger, integrated with an adaptive frontier assigner for optimal coverage balancing the exploration and reconstruction objectives. In extensive ROS-Gazebo simulations, SPACE demonstrated superior performance over state-of-the-art approaches in both exploration and mapping metrics.

Figures (7)

• Figure 1: The Ghosting Trail Problem: Formation of ghosting regions & poor quality spatial maps due to inter-robot visibility during mapping and exploration. (a), (b) and (c), (d) represents explored and composite (where each color represents a local map by each robot) 3D map using RRT with Kimera-Multi tian2022kimera and SPACE with RTABMap labbe2019rtab, respectively.
• Figure 2: Overview of the proposed methodology. The Blue-shaded components are Mapping-related modules, and the Orange-shaded components are Exploration-realated modules. Marked with an asterisk(*) are novel elements introduced in this paper for multi-robot exploration.
• Figure 3: (Left) Contour-based Frontier Detection keidar2014efficient (Center) Bi-Variate Spatial Frontier Detection of SPACE (Right) Grid map with Translated Spatial Frontiers.
• Figure 4: Performance of the spatial mapping in multi-robot exploration. Here, we used the SPACE exploration approach described in Sec. \ref{['sec:methodology']} in all mapping variants for a fair comparison. The top row subplots show the 3D reconstruction accuracy in the three scenarios tested ((Left): AWS House World (3 Robots), (Middle): AWS Bookstore World (3 Robots), and (Right): AWS Bookstore World (6 Robots)). In the bottom row, the (Left) plot shows the detailed robot-wise performance variations, and (Right) plot shows the impact of the ghosting trail effect (bounding volume of the reconstruction inaccuracies) by increasing the robot density in a given area. SPACE-RTABMap is invariant to the ghosting effect with an increase in density.
• Figure 5: Performance Analysis of Exploration Strategies with respect to Total Spatial Coverage in tested environments (Left) AWS House World (3 robots) (Right) AWS Bookstore World (3 robots).