Deep Learning Traversability Estimator for Mobile Robots in Unstructured Environments

TL;DR

This paper introduces a deep learning framework for real-time terrain traversability estimation in unstructured environments, demonstrating high accuracy and efficiency in simulation and successful transfer to real-world Martian terrain data.

Contribution

The work presents an end-to-end deep learning approach trained on synthetic data and successfully transferred to real terrain maps, improving real-time robotic navigation.

Findings

Achieves over 94% recall at 30% of the computational time in simulation.

Successfully transfers and fine-tunes the model on real Martian terrain data.

Outperforms models trained solely on limited real data.

Abstract

Terrain traversability analysis plays a major role in ensuring safe robotic navigation in unstructured environments. However, real-time constraints frequently limit the accuracy of online tests especially in scenarios where realistic robot-terrain interactions are complex to model. In this context, we propose a deep learning framework trained in an end-to-end fashion from elevation maps and trajectories to estimate the occurrence of failure events. The network is first trained and tested in simulation over synthetic maps generated by the OpenSimplex algorithm. The prediction performance of the Deep Learning framework is illustrated by being able to retain over 94% recall of the original simulator at 30% of the computational time. Finally, the network is transferred and tested on real elevation maps collected by the SEEKER consortium during the Martian rover test trial in the Atacama desert in Chile. We show that transferring and fine-tuning of an application-independent pre-trained model retains better performance than training uniquely on scarcely available real data.