From Underground Mines to Offices: A Versatile and Robust Framework for Range-Inertial SLAM

Lorenzo Montano-Oliván,Julio A. Placed,Luis Montano,María T. Lázaro

2024-07-20

Abstract:Simultaneous Localization and Mapping (SLAM) is an essential component of autonomous robotic applications and self-driving vehicles, enabling them to understand and operate in their environment. Many SLAM systems have been proposed in the last decade, but they are often complex to adapt to different settings or sensor setups. In this work, we present LiDAR Graph-SLAM (LG-SLAM), a versatile range-inertial SLAM framework that can be adapted to different types of sensors and environments, from underground mines to offices with minimal parameter tuning. Our system integrates range, inertial and GNSS measurements into a graph-based optimization framework. We also use a refined submap management approach and a robust loop closure method that effectively accounts for uncertainty in the identification and validation of putative loop closures, ensuring global consistency and robustness. Enabled by a parallelized architecture and GPU integration, our system achieves pose estimation at LiDAR frame rate, along with online loop closing and graph optimization. We validate our system in diverse environments using public datasets and real-world data, consistently achieving an average error below 20 cm and outperforming other state-of-the-art algorithms.

Robotics

What problem does this paper attempt to address?

The paper aims to address the problem of precise mapping and localization for autonomous navigation of robots and self-driving cars in different environments. Specifically, the authors propose a framework called LiDAR Graph-SLAM (LG-SLAM), which is a simultaneous localization and mapping (SLAM) system suitable for various sensor configurations and environmental conditions. The system has the following features:

1. **High Adaptability**: LG-SLAM can adapt to different types of environments, from underground mines to offices, and can work with multiple sensors (such as LiDAR, IMU, and GNSS) without the need for complex parameter adjustments.

2. **Good Robustness**: The system combines range, inertial, and global navigation satellite system (GNSS) measurements to achieve precise pose estimation through a graph-based optimization framework. Additionally, it employs an improved submap management method and an efficient loop closure detection mechanism to ensure global consistency.

3. **Real-time Performance**: Thanks to its parallel architecture and GPU acceleration technology, the system can achieve pose estimation at LiDAR frame rates and perform loop closure detection and graph optimization online.

4. **High Accuracy**: Experimental results show that LG-SLAM performs excellently in various public datasets and real-world scenarios, with an average error of less than 20 centimeters, outperforming other existing algorithms.

With these features, LG-SLAM can achieve efficient and accurate map construction and robot localization in various complex environments, particularly excelling in conditions where visual sensors struggle to operate.