--The advancement of robotics and autonomous navigation systems hinges on the ability to accurately predict terrain traversability. Traditional methods for generating datasets to train these prediction models often involve putting robots into potentially hazardous environments, posing risks to equipment and safety. T o solve this problem, we present ZeST, a novel approach leveraging visual reasoning capabilities of Large Language Models (LLMs) to create a traversability map in real-time without exposing robots to danger . Our approach not only performs zero-shot traversability and mitigates the risks associated with real-world data collection but also accelerates the development of advanced navigation systems, offering a cost-effective and scalable solution. T o support our findings, we present navigation results, in both controlled indoor and unstructured outdoor environments. As shown in the experiments, our method provides safer navigation when compared to other state-of-the-art methods, constantly reaching the final goal. The development of autonomous navigation systems is a cornerstone of robotics, with terrain traversability prediction being a critical component [1], [2], [3], [4], [5]. Traversability prediction refers to the ability of a robot to assess whether a given terrain is passable or poses risks to its operation.

Reliable navigation in unstructured, real-world environments remains a significant challenge for embodied agents, especially when operating across diverse terrains, weather conditions, and sensor configurations. In this paper, we introduce GeNIE (Generalizable Navigation System for In-the-Wild Environments), a robust navigation framework designed for global deployment. GeNIE integrates a generalizable traversability prediction model built on SAM2 with a novel path fusion strategy that enhances planning stability in noisy and ambiguous settings. We deployed GeNIE in the Earth Rover Challenge (ERC) at ICRA 2025, where it was evaluated across six countries spanning three continents. GeNIE took first place and achieved 79% of the maximum possible score, outperforming the second-best team by 17%, and completed the entire competition without a single human intervention. These results set a new benchmark for robust, generalizable outdoor robot navigation. We will release the codebase, pretrained model weights, and newly curated datasets to support future research in real-world navigation.

The increasing use of robots in unstructured environments necessitates the development of effective perception and navigation strategies to enable field robots to successfully perform their tasks. In particular, it is key for such robots to understand where in their environment they can and cannot travel -- a task known as traversability estimation. However, existing geometric approaches to traversability estimation may fail to capture nuanced representations of traversability, whereas vision-based approaches typically either involve manually annotating a large number of images or require robot experience. In addition, existing methods can struggle to address domain shifts as they typically do not learn during deployment. To this end, we propose a human-in-the-loop (HiL) method for traversability estimation that prompts a human for annotations as-needed. Our method uses a foundation model to enable rapid learning on new annotations and to provide accurate predictions even when trained on a small number of quickly-provided HiL annotations. We extensively validate our method in simulation and on real-world data, and demonstrate that it can provide state-of-the-art traversability prediction performance.

Abstract-- We propose WildFusion, a novel approach for 3D scene reconstruction in unstructured, in-the-wild environments using multimodal implicit neural representations. This multimodal fusion generates comprehensive, continuous environmental representations, including pixel-level geometry, color, semantics, and traversability. Through real-world experiments on legged robot navigation in challenging forest environments, WildFusion demonstrates improved route selection by accurately predicting traversability. Our results highlight its potential to advance robotic navigation and 3D mapping in complex outdoor terrains. Robots need effective environmental representations to navigate safely and accomplish tasks successfully in unstructured Figure 1: WildFusion integrates LiDAR, camera, microphones, outdoor environments - often referred to as "in-thewild" and tactile sensors with implicit neural representations for settings such as monitoring high-voltage power lines continuous 3D scene reconstruction.

Traversability assessment of deformable terrain is vital for safe rover navigation on planetary surfaces. Machine learning (ML) is a powerful tool for traversability prediction but faces predictive uncertainty. This uncertainty leads to prediction errors, increasing the risk of wheel slips and immobilization for planetary rovers. To address this issue, we integrate principal approaches to uncertainty handling -- quantification, exploitation, and adaptation -- into a single learning and planning framework for rover navigation. The key concept is \emph{deep probabilistic traversability}, forming the basis of an end-to-end probabilistic ML model that predicts slip distributions directly from rover traverse observations. This probabilistic model quantifies uncertainties in slip prediction and exploits them as traversability costs in path planning. Its end-to-end nature also allows adaptation of pre-trained models with in-situ traverse experience to reduce uncertainties. We perform extensive simulations in synthetic environments that pose representative uncertainties in planetary analog terrains. Experimental results show that our method achieves more robust path planning under novel environmental conditions than existing approaches.

Successful deployment of mobile robots in unstructured domains requires an understanding of the environment and terrain to avoid hazardous areas, getting stuck, and colliding with obstacles. Traversability estimation--which predicts where in the environment a robot can travel--is one prominent approach that tackles this problem. Existing geometric methods may ignore important semantic considerations, while semantic segmentation approaches involve a tedious labeling process. Recent self-supervised methods reduce labeling tedium, but require additional data or models and tend to struggle to explicitly label untraversable areas. To address these limitations, we introduce a weakly-supervised method for relative traversability estimation. Our method involves manually annotating the relative traversability of a small number of point pairs, which significantly reduces labeling effort compared to traditional segmentation-based methods and avoids the limitations of self-supervised methods. We further improve the performance of our method through a novel cross-image labeling strategy and loss function. We demonstrate the viability and performance of our method through deployment on a mobile robot in outdoor environments.

As robotic navigation techniques in perception and planning advance, mobile robots increasingly venture into off-road environments involving complex traversability. However, selecting suitable planning methods remains a challenge due to their algorithmic diversity, as each offers unique benefits. To aid in algorithm design, we introduce BenchNav, an open-source PyTorch-based simulation platform for benchmarking off-road navigation with uncertain traversability. Built upon Gymnasium, BenchNav provides three key features: 1) a data generation pipeline for preparing synthetic natural environments, 2) built-in machine learning models for traversability prediction, and 3) consistent execution of path and motion planning across different algorithms. We show BenchNav's versatility through simulation examples in off-road environments, employing three representative planning algorithms from different domains. https://github.com/masafumiendo/benchnav

Natural environments such as forests and grasslands are challenging for robotic navigation because of the false perception of rigid obstacles from high grass, twigs, or bushes. In this work, we present Wild Visual Navigation (WVN), an online self-supervised learning system for visual traversability estimation. The system is able to continuously adapt from a short human demonstration in the field, only using onboard sensing and computing. One of the key ideas to achieve this is the use of high-dimensional features from pre-trained self-supervised models, which implicitly encode semantic information that massively simplifies the learning task. Further, the development of an online scheme for supervision generator enables concurrent training and inference of the learned model in the wild. We demonstrate our approach through diverse real-world deployments in forests, parks, and grasslands. Our system is able to bootstrap the traversable terrain segmentation in less than 5 min of in-field training time, enabling the robot to navigate in complex, previously unseen outdoor terrains. Code: https://bit.ly/498b0CV - Project page:https://bit.ly/3M6nMHH

Accurate and robust navigation in unstructured environments requires fusing data from multiple sensors. Such fusion ensures that the robot is better aware of its surroundings, including areas of the environment that are not immediately visible, but were visible at a different time. To solve this problem, we propose a method for traversability prediction in challenging outdoor environments using a sequence of RGB and depth images fused with pose estimations. Our method, termed WayFASTER (Waypoints-Free Autonomous System for Traversability with Enhanced Robustness), uses experience data recorded from a receding horizon estimator to train a self-supervised neural network for traversability prediction, eliminating the need for heuristics. Our experiments demonstrate that our method excels at avoiding geometric obstacles, and correctly detects that traversable terrains, such as tall grass, can be navigable. By using a sequence of images, WayFASTER significantly enhances the robot's awareness of its surroundings, enabling it to predict the traversability of terrains that are not immediately visible. This enhanced awareness contributes to better navigation performance in environments where such predictive capabilities are essential.

Reliable estimation of terrain traversability is critical for the successful deployment of autonomous systems in wild, outdoor environments. Given the lack of large-scale annotated datasets for off-road navigation, strictly-supervised learning approaches remain limited in their generalization ability. To this end, we introduce a novel, image-based self-supervised learning method for traversability prediction, leveraging a state-of-the-art vision foundation model for improved out-of-distribution performance. Our method employs contrastive representation learning using both human driving data and instance-based segmentation masks during training. We show that this simple, yet effective, technique drastically outperforms recent methods in predicting traversability for both on- and off-trail driving scenarios. We compare our method with recent baselines on both a common benchmark as well as our own datasets, covering a diverse range of outdoor environments and varied terrain types. We also demonstrate the compatibility of resulting costmap predictions with a model-predictive controller. Finally, we evaluate our approach on zero- and few-shot tasks, demonstrating unprecedented performance for generalization to new environments. Videos and additional material can be found here: \url{https://sites.google.com/view/visual-traversability-learning}.