Ali Agha

Xianmei Lei, Taeyeon Kim, Nicolas Marchal et al.

Semantic object mapping in uncertain, perceptually degraded environments during long-range multi-robot autonomous exploration tasks such as search-and-rescue is important and challenging. During such missions, high recall is desirable to avoid missing true target objects and high precision is also critical to avoid wasting valuable operational time on false positives. Given recent advancements in visual perception algorithms, the former is largely solvable autonomously, but the latter is difficult to address without the supervision of a human operator. However, operational constraints such as mission time, computational requirements, mesh network bandwidth and so on, can make the operator's task infeasible unless properly managed. We propose the Early Recall, Late Precision (EaRLaP) semantic object mapping pipeline to solve this problem. EaRLaP was used by Team CoSTAR in DARPA Subterranean Challenge, where it successfully detected all the artifacts encountered by the team of robots. We will discuss these results and performance of the EaRLaP on various datasets.

Jonas Frey, Manthan Patel, Deegan Atha et al.

Autonomous navigation at high speeds in off-road environments necessitates robots to comprehensively understand their surroundings using onboard sensing only. The extreme conditions posed by the off-road setting can cause degraded camera image quality due to poor lighting and motion blur, as well as limited sparse geometric information available from LiDAR sensing when driving at high speeds. In this work, we present RoadRunner, a novel framework capable of predicting terrain traversability and an elevation map directly from camera and LiDAR sensor inputs. RoadRunner enables reliable autonomous navigation, by fusing sensory information, handling of uncertainty, and generation of contextually informed predictions about the geometry and traversability of the terrain while operating at low latency. In contrast to existing methods relying on classifying handcrafted semantic classes and using heuristics to predict traversability costs, our method is trained end-to-end in a self-supervised fashion. The RoadRunner network architecture builds upon popular sensor fusion network architectures from the autonomous driving domain, which embed LiDAR and camera information into a common Bird's Eye View perspective. Training is enabled by utilizing an existing traversability estimation stack to generate training data in hindsight in a scalable manner from real-world off-road driving datasets. Furthermore, RoadRunner improves the system latency by a factor of roughly 4, from 500 ms to 140 ms, while improving the accuracy for traversability costs and elevation map predictions. We demonstrate the effectiveness of RoadRunner in enabling safe and reliable off-road navigation at high speeds in multiple real-world driving scenarios through unstructured desert environments.

Chanyoung Chung, Georgios Georgakis, Patrick Spieler et al.

Understanding terrain topology at long-range is crucial for the success of off-road robotic missions, especially when navigating at high-speeds. LiDAR sensors, which are currently heavily relied upon for geometric mapping, provide sparse measurements when mapping at greater distances. To address this challenge, we present a novel learning-based approach capable of predicting terrain elevation maps at long-range using only onboard egocentric images in real-time. Our proposed method is comprised of three main elements. First, a transformer-based encoder is introduced that learns cross-view associations between the egocentric views and prior bird-eye-view elevation map predictions. Second, an orientation-aware positional encoding is proposed to incorporate the 3D vehicle pose information over complex unstructured terrain with multi-view visual image features. Lastly, a history-augmented learn-able map embedding is proposed to achieve better temporal consistency between elevation map predictions to facilitate the downstream navigational tasks. We experimentally validate the applicability of our proposed approach for autonomous offroad robotic navigation in complex and unstructured terrain using real-world offroad driving data. Furthermore, the method is qualitatively and quantitatively compared against the current state-of-the-art methods. Extensive field experiments demonstrate that our method surpasses baseline models in accurately predicting terrain elevation while effectively capturing the overall terrain topology at long-ranges. Finally, ablation studies are conducted to highlight and understand the effect of key components of the proposed approach and validate their suitability to improve offroad robotic navigation capabilities.

Antonello Longo, Chanyoung Chung, Matteo Palieri et al.

Autonomous robots are increasingly playing key roles as support platforms for human operators in high-risk, dangerous applications. To accomplish challenging tasks, an efficient human-robot cooperation and understanding is required. While typically robotic planning leverages 3D geometric information, human operators are accustomed to a high-level compact representation of the environment, like top-down 2D maps representing the Building Information Model (BIM). 3D scene graphs have emerged as a powerful tool to bridge the gap between human readable 2D BIM and the robot 3D maps. In this work, we introduce Pixels-to-Graph (Pix2G), a novel lightweight method to generate structured scene graphs from image pixels and LiDAR maps in real-time for the autonomous exploration of unknown environments on resource-constrained robot platforms. To satisfy onboard compute constraints, the framework is designed to perform all operation on CPU only. The method output are a de-noised 2D top-down environment map and a structure-segmented 3D pointcloud which are seamlessly connected using a multi-layer graph abstracting information from object-level up to the building-level. The proposed method is quantitatively and qualitatively evaluated during real-world experiments performed using the NASA JPL NeBula-Spot legged robot to autonomously explore and map cluttered garage and urban office like environments in real-time.

Jason Jabbour, Dong-Ki Kim, Max Smith et al.

Vision-Language-Action (VLA) models have advanced robotic capabilities but remain challenging to deploy on resource-limited hardware. Pruning has enabled efficient compression of large language models (LLMs), yet it is largely understudied in robotics. Surprisingly, we observe that pruning VLA models leads to drastic degradation and increased safety violations. We introduce GLUESTICK, a post-pruning recovery method that restores much of the original model's functionality while retaining sparsity benefits. Our method performs a one-time interpolation between the dense and pruned models in weight-space to compute a corrective term. This correction is used during inference by each pruned layer to recover lost capabilities with minimal overhead. GLUESTICK requires no additional training, is agnostic to the pruning algorithm, and introduces a single hyperparameter that controls the tradeoff between efficiency and accuracy. Across diverse VLA architectures and tasks in manipulation and navigation, GLUESTICK achieves competitive memory efficiency while substantially recovering success rates and reducing safety violations. Additional material can be found at: https://gluestick-vla.github.io/.

Yutian Chen, Yuheng Qiu, Ruogu Li et al.

We propose Confidence-Guided Token Merging (Co-Me), an acceleration mechanism for visual geometric transformers without retraining or finetuning the base model. Co-Me distilled a light-weight confidence predictor to rank tokens by uncertainty and selectively merge low-confidence ones, effectively reducing computation while maintaining spatial coverage. Compared to similarity-based merging or pruning, the confidence signal in Co-Me reliably indicates regions emphasized by the transformer, enabling substantial acceleration without degrading performance. Co-Me applies seamlessly to various multi-view and streaming visual geometric transformers, achieving speedups that scale with sequence length. When applied to VGGT and MapAnything, Co-Me achieves up to $11.3\times$ and $7.2\times$ speedup, making visual geometric transformers practical for real-time 3D perception and reconstruction.

Arthur Zhang, Xiangyun Meng, Luca Calliari et al.

Robots must adapt to diverse human instructions and operate safely in unstructured, open-world environments. Recent Vision-Language models (VLMs) offer strong priors for grounding language and perception, but remain difficult to steer for navigation due to differences in action spaces and pretraining objectives that hamper transferability to robotics tasks. Towards addressing this, we introduce VENTURA, a vision-language navigation system that finetunes internet-pretrained image diffusion models for path planning. Instead of directly predicting low-level actions, VENTURA generates a path mask (i.e. a visual plan) in image space that captures fine-grained, context-aware navigation behaviors. A lightweight behavior-cloning policy grounds these visual plans into executable trajectories, yielding an interface that follows natural language instructions to generate diverse robot behaviors. To scale training, we supervise on path masks derived from self-supervised tracking models paired with VLM-augmented captions, avoiding manual pixel-level annotation or highly engineered data collection setups. In extensive real-world evaluations, VENTURA outperforms state-of-the-art foundation model baselines on object reaching, obstacle avoidance, and terrain preference tasks, improving success rates by 33% and reducing collisions by 54% across both seen and unseen scenarios. Notably, we find that VENTURA generalizes to unseen combinations of distinct tasks, revealing emergent compositional capabilities. Videos, code, and additional materials: https://venturapath.github.io

Ali Agha, Kyohei Otsu, Benjamin Morrell et al.

This paper presents and discusses algorithms, hardware, and software architecture developed by the TEAM CoSTAR (Collaborative SubTerranean Autonomous Robots), competing in the DARPA Subterranean Challenge. Specifically, it presents the techniques utilized within the Tunnel (2019) and Urban (2020) competitions, where CoSTAR achieved 2nd and 1st place, respectively. We also discuss CoSTAR's demonstrations in Martian-analog surface and subsurface (lava tubes) exploration. The paper introduces our autonomy solution, referred to as NeBula (Networked Belief-aware Perceptual Autonomy). NeBula is an uncertainty-aware framework that aims at enabling resilient and modular autonomy solutions by performing reasoning and decision making in the belief space (space of probability distributions over the robot and world states). We discuss various components of the NeBula framework, including: (i) geometric and semantic environment mapping; (ii) a multi-modal positioning system; (iii) traversability analysis and local planning; (iv) global motion planning and exploration behavior; (i) risk-aware mission planning; (vi) networking and decentralized reasoning; and (vii) learning-enabled adaptation. We discuss the performance of NeBula on several robot types (e.g. wheeled, legged, flying), in various environments. We discuss the specific results and lessons learned from fielding this solution in the challenging courses of the DARPA Subterranean Challenge competition.

E. Natasha Stavros, Ali Agha, Allen Sirota et al.

Wildfire has had increasing impacts on society as the climate changes and the wildland urban interface grows. As such, there is a demand for innovative solutions to help manage fire. Managing wildfire can include proactive fire management such as prescribed burning within constrained areas or advancements for reactive fire management (e.g., fire suppression). Because of the growing societal impact, the JPL BlueSky program sought to assess the current state of fire management and technology and determine areas with high return on investment. To accomplish this, we met with the national interagency Unmanned Aerial System (UAS) Advisory Group (UASAG) and with leading technology transfer experts for fire science and management applications. We provide an overview of the current state as well as an analysis of the impact, maturity and feasibility of integrating different technologies that can be developed by JPL. Based on the findings, the highest return on investment technologies for fire management are first to develop single micro-aerial vehicle (MAV) autonomy, autonomous sensing over fire, and the associated data and information system for active fire local environment mapping. Once this is completed for a single MAV, expanding the work to include many in a swarm would require further investment of distributed MAV autonomy and MAV swarm mechanics, but could greatly expand the breadth of application over large fires. Important to investing in these technologies will be in developing collaborations with the key influencers and champions for using UAS technology in fire management.