Multi sensor dataset improves robot navigation in off road terrain
MS-DOTT: A Multi-Modal Dataset for Off-Road Traversability Estimation
Robotics
Summary
Off-road environments are hard for robots to navigate because the ground and lighting can change a lot. The authors created a new dataset called MS-DOTT that combines data from multiple sensors, including 3D imaging, radar, and vehicle signals, to help robots understand the terrain better. The dataset includes recordings in different light conditions and captures both detailed nearby terrain and faraway environment. They tested the dataset by using it with a modern method to estimate how easy it is for a robot to cross different parts of off-road terrain. This resource aims to help improve autonomous robots working in difficult outdoor areas.
What this means in practice
- •For agriculture robotics teams: Improve robot navigation across rugged farmland by using multi-sensor data to better estimate terrain difficulty in varied lighting conditions.
- •For search and rescue operators: Enhance off-road robot perception to help safely navigate unknown disaster sites using integrated sensor data for accurate terrain assessment.
Authors
Shivani Subramanyan, Saksham Sharma, Samahith SA, Ahaan Kotian, Suresh Sundaram
Abstract
Autonomous robots are gaining traction in off-road environments, driven by their rising demand in agriculture, defense, and search-and-rescue operations. However, these environments are inherently unpredictable and navigation remains a formidable challenge given the complex environmental conditions including challenging terrain and varying illumination. Off-road autonomy requires robust perception across multiple spatial scales and imaging modalities, yet most sensing setups struggle to capture both far-field environmental context and near-field terrain detail simultaneously across various illumination conditions. To facilitate development of robust off-road perception and traversability estimation systems, we propose MS-DOTT (Multi-Sensor Dataset for Off-road Terrain Traversability) a comprehensive multi-modal off-road dataset with long-range 3D sensing and short-range, fine-grained perception to facilitate high-fidelity terrain understanding by combining proprioceptive signals for tracking vehicle dynamics with exteroceptive imaging and pointcloud data. The dataset includes trajectories recorded across nominal lighting, low-light, and no-light conditions with diverse 2D and 3D sensing for both long-range large-scale and short-range high-resolution sensing, as well as 4D RADAR measurements that provide velocity estimates for detected pointclouds. We demonstrate the utility of the dataset by adapting a SOTA multi-sensor fusion method for the purpose of traversability estimation. This dataset will aid the robotics community in advancing research and development of methods for traversability estimation, off-road perception and multi-sensor fusion in unstructured environments.