Papers

Improving robotic navigation is critical for extending exploration range and enhancing operational efficiency. Vision-based navigation relying on traditional CCD or CMOS cameras faces major challenges when complex illumination conditions are paired with motion, limiting the range and accessibility of mobile planetary robots. In this study, we propose a novel approach to planetary navigation that leverages the unique imaging capabilities of Single-Photon Avalanche Diode (SPAD) cameras. We present the first comprehensive evaluation of single-photon imaging as an alternative passive sensing technology for robotic exploration missions targeting perceptually challenging locations, with a special emphasis on high-latitude lunar regions. We detail the operating principles and performance characteristics of SPAD cameras, assess their advantages and limitations in addressing key perception challenges of upcoming exploration missions to the Moon, and benchmark their performance under representative illumination conditions.

Sampling-based trajectory planners are widely used for agile autonomous driving due to their ability to generate fast, smooth, and kinodynamically feasible trajectories. However, their behavior is often governed by a cost function with manually tuned, static weights, which forces a tactical compromise that is suboptimal across the wide range of scenarios encountered in a race. To address this shortcoming, we propose using a Reinforcement Learning (RL) agent as a high-level behavioral selector that dynamically switches the cost function parameters of an analytical, low-level trajectory planner during runtime. We show the effectiveness of our approach in simulation in an autonomous racing environment where our RL-based planner achieved 0% collision rate while reducing overtaking time by up to 60% compared to state-of-the-art static planners. Our new agent now dynamically switches between aggressive and conservative behaviors, enabling interactive maneuvers unattainable with static configurations. These results demonstrate that integrating reinforcement learning as a high-level selector resolves the inherent trade-off between safety and competitiveness in autonomous racing planners. The proposed methodology offers a pathway toward adaptive yet interpretable motion planning for broader autonomous driving applications.

Four-channel bilateral control is a method for achieving remote control with force feedback and adjustment operability by synchronizing the positions and forces of two manipulators. This is expected to significantly improve the operability of the remote control in contact-rich tasks. Among these, 4-channel bilateral control on the Cartesian coordinate system is advantageous owing to its suitability for manipulators with different structures and because it allows the dynamics in the Cartesian coordinate system to be adjusted by adjusting the control parameters, thus achieving intuitive operability for humans. This paper proposes a 4-channel bilateral control method that achieves the desired dynamics by decoupling each dimension in the Cartesian coordinate system regardless of the scaling factor.

As the global reach of online higher education continues to grow, universities are increasingly accommodating students from diverse cultural backgrounds (Tereshko et al., 2024). This can present a number of challenges including linguistic barriers (Ullah et al., 2021), cultural differences in learning style (Omidvar & Tan, 2012), cultural sensitivity in course design (Nguyen, 2022) and perceived isolation when students feel their perspectives or experiences are not reflected or valued in the learning environment (Hansen-Brown et al., 2022). Ensuring active engagement and reasonable learning outcomes in such a environments requires distance educational systems that are not only adaptive but also culturally resonant (Dalle et al., 2024). Both embodied and virtual AI-Agents have great potential in this regard as they can facilitate personalized learning and adapt their interactions and content delivery to align with students' cultural context. In addition, Generative AI (GAI), such as, Large Language Models (LLMs) can amplify the potential for these culturally aware AI agents to address educational challenges due to their advanced capacity for understanding and generating contextually relevant content (Wang et al., 2024). This chapter reviews existing research and suggests the usage of culturally aware AI-Agents, powered by GAI, to foster engagement and improve learning outcomes in culturally diverse online higher education environments.

Energy-efficient and high-performance motor control remains a critical challenge in robotics, particularly for high-dimensional continuous control tasks with limited onboard resources. While Deep Reinforcement Learning (DRL) has achieved remarkable results, its computational demands and energy consumption limit deployment in resource-constrained environments. This paper introduces a novel framework combining population-coded Spiking Neural Networks (SNNs) with DRL to address these challenges. Our approach leverages the event-driven, asynchronous computation of SNNs alongside the robust policy optimization capabilities of DRL, achieving a balance between energy efficiency and control performance. Central to this framework is the Population-coded Spiking Actor Network (PopSAN), which encodes high-dimensional observations into neuronal population activities and enables optimal policy learning through gradient-based updates. We evaluate our method on the Isaac Gym platform using the PixMC benchmark with complex robotic manipulation tasks. Experimental results on the Franka robotic arm demonstrate that our approach achieves energy savings of up to 96.10% compared to traditional Artificial Neural Networks (ANNs) while maintaining comparable control performance. The trained SNN policies exhibit robust finger position tracking with minimal deviation from commanded trajectories and stable target height maintenance during pick-and-place operations. These results position population-coded SNNs as a promising solution for energy-efficient, high-performance robotic control in resource-constrained applications, paving the way for scalable deployment in real-world robotics systems.

Efficient exploration and mapping in unknown indoor environments is a fundamental challenge, with high stakes in time-critical settings. In current systems, robot perception remains confined to line-of-sight; occluded regions remain unknown until physically traversed, leading to inefficient exploration when layouts deviate from prior assumptions. In this work, we bring non-line-of-sight (NLOS) sensing to robotic exploration. We leverage single-photon LiDARs, which capture time-of-flight histograms that encode the presence of hidden objects - allowing robots to look around blind corners. Recent single-photon LiDARs have become practical and portable, enabling deployment beyond controlled lab settings. Prior NLOS works target 3D reconstruction in static, lab-based scenarios, and initial efforts toward NLOS-aided navigation consider simplified geometries. We introduce SuperEx, a framework that integrates NLOS sensing directly into the mapping-exploration loop. SuperEx augments global map prediction with beyond-line-of-sight cues by (i) carving empty NLOS regions from timing histograms and (ii) reconstructing occupied structure via a two-step physics-based and data-driven approach that leverages structural regularities. Evaluations on complex simulated maps and the real-world KTH Floorplan dataset show a 12% gain in mapping accuracy under < 30% coverage and improved exploration efficiency compared to line-of-sight baselines, opening a path to reliable mapping beyond direct visibility.

Motion planning in complex scenarios is a core challenge in autonomous driving. Conventional methods apply predefined rules or learn from driving data to generate trajectories, while recent approaches leverage large language models (LLMs) for decision-making. However, it remains unclear whether LLMs truly capture human driving logic. We propose Align2Act, a motion planning framework that transforms instruction-tuned LLMs into interpretable planners aligned with human behavior. We derive structured driving instructions based on human reasoning patterns (e.g., anticipate hazards, yield at intersections) and traffic rules (e.g., stop at red lights, maintain lane boundaries). Our Align2ActChain module guides step-by-step reasoning to produce both an interpretable rationale and a safe trajectory. By fine-tuning LLaMA-2-7B with LoRA on one million scenarios from the nuPlan dataset, our method achieves an open-loop score of 85.17 and closed-loop scores of 70.31 (non-reactive) and 66.96 (reactive) on Test14-random. Unlike prior work focused on synthetic or open-loop settings, we demonstrate improved planning quality and human-likeness on the real-world nuPlan closed-loop benchmark. Ablation studies confirm that structured reasoning significantly improves performance over baseline LLM planners.

Galilean symmetry is the natural symmetry of inertial motion that underpins Newtonian physics. Although rigid-body symmetry is one of the most established and fundamental tools in robotics, there appears to be no comparable treatment of Galilean symmetry for a robotics audience. In this paper, we present a robotics-tailored exposition of Galilean symmetry that leverages the community's familiarity with and understanding of rigid-body transformations and pose representations. Our approach contrasts with common treatments in the physics literature that introduce Galilean symmetry as a stepping stone to Einstein's relativity. A key insight is that the Galilean matrix Lie group can be used to describe two different pose representations, Galilean frames, that use inertial velocity in the state definition, and extended poses, that use coordinate velocity. We provide three examples where applying the Galilean matrix Lie-group algebra to robotics problems is straightforward and yields significant insights: inertial navigation above the rotating Earth, manipulator kinematics, and sensor data fusion under temporal uncertainty. We believe that the time is right for the robotics community to benefit from rediscovering and extending this classical material and applying it to modern problems.

Quadrupedal robots exhibit a wide range of viable gaits, but generating specific footfall sequences often requires laborious expert tuning of numerous variables, such as touch-down and lift-off events and holonomic constraints for each leg. This paper presents a unified reinforcement learning framework for generating versatile quadrupedal gaits by leveraging the intrinsic symmetries and velocity-period relationship of dynamic legged systems. We propose a symmetry-guided reward function design that incorporates temporal, morphological, and time-reversal symmetries. By focusing on preserved symmetries and natural dynamics, our approach eliminates the need for predefined trajectories, enabling smooth transitions between diverse locomotion patterns such as trotting, bounding, half-bounding, and galloping. Implemented on the Unitree Go2 robot, our method demonstrates robust performance across a range of speeds in both simulations and hardware tests, significantly improving gait adaptability without extensive reward tuning or explicit foot placement control. This work provides insights into dynamic locomotion strategies and underscores the crucial role of symmetries in robotic gait design.

We propose MonoSE(3)-Diffusion, a monocular SE(3) diffusion framework that formulates markerless, image-based robot pose estimation as a conditional denoising diffusion process. The framework consists of two processes: a visibility-constrained diffusion process for diverse pose augmentation and a timestep-aware reverse process for progressive pose refinement. The diffusion process progressively perturbs ground-truth poses to noisy transformations for training a pose denoising network. Importantly, we integrate visibility constraints into the process, ensuring the transformations remain within the camera field of view. Compared to the fixed-scale perturbations used in current methods, the diffusion process generates in-view and diverse training poses, thereby improving the network generalization capability. Furthermore, the reverse process iteratively predicts the poses by the denoising network and refines pose estimates by sampling from the diffusion posterior of current timestep, following a scheduled coarse-to-fine procedure. Moreover, the timestep indicates the transformation scales, which guide the denoising network to achieve more accurate pose predictions. The reverse process demonstrates higher robustness than direct prediction, benefiting from its timestep-aware refinement scheme. Our approach demonstrates improvements across two benchmarks (DREAM and RoboKeyGen), achieving a notable AUC of 66.75 on the most challenging dataset, representing a 32.3% gain over the state-of-the-art.

Achieving persistent tracking of multiple dynamic targets over a large spatial area poses significant challenges for a single-robot system with constrained sensing capabilities. As the robot moves to track different targets, the ones outside the field of view accumulate uncertainty, making them progressively harder to track. An effective path planning algorithm must manage uncertainty over a long horizon and account for the risk of permanently losing track of targets that remain unseen for too long. However, most existing approaches rely on short planning horizons and assume small, bounded environments, resulting in poor tracking performance and target loss in large-scale scenarios. In this paper, we present a hierarchical planner for tracking multiple moving targets with an aerial vehicle. To address the challenge of tracking non-static targets, our method incorporates motion models and uncertainty propagation during path execution, allowing for more informed decision-making. We decompose the multi-target tracking task into sub-tasks of single target search and detection, and our proposed pipeline consists a novel low-level coverage planner that enables searching for a target in an evolving belief area, and an estimation method to assess the likelihood of success for each sub-task, making it possible to convert the active target tracking task to a Markov decision process (MDP) that we solve with a tree-based algorithm to determine the sequence of sub-tasks. We validate our approach in simulation, demonstrating its effectiveness compared to existing planners for active target tracking tasks, and our proposed planner outperforms existing approaches, achieving a reduction of 11-70% in final uncertainty across different environments.

This paper presents MicroRoboScope, a portable, compact, and versatile microrobotic experimentation platform designed for real-time, closed-loop control of both magnetic and acoustic microrobots. The system integrates an embedded computer, microscope, power supplies, and control circuitry into a single, low-cost and fully integrated apparatus. Custom control software developed in Python and Arduino C++ handles live video acquisition, microrobot tracking, and generation of control signals for electromagnetic coils and acoustic transducers. The platform's multi-modal actuation, accessibility, and portability make it suitable not only for specialized research laboratories but also for educational and outreach settings. By lowering the barrier to entry for microrobotic experimentation, this system enables new opportunities for research, education, and translational applications in biomedicine, tissue engineering, and robotics.

Coordinating heterogeneous robot fleets to achieve multiple goals is challenging in multi-robot systems. We introduce an open-source and extensible framework for centralized multi-robot task planning and scheduling that leverages LLMs to enable fleets of heterogeneous robots to accomplish multiple tasks. RobotFleet provides abstractions for planning, scheduling, and execution across robots deployed as containerized services to simplify fleet scaling and management. The framework maintains a shared declarative world state and two-way communication for task execution and replanning. By modularizing each layer of the autonomy stack and using LLMs for open-world reasoning, RobotFleet lowers the barrier to building scalable multi-robot systems. The code can be found here: https://github.com/therohangupta/robot-fleet.

Reinforcement learning (RL) has become a foundational approach for enabling intelligent robotic behavior in dynamic and uncertain environments. This work presents an in-depth review of RL principles, advanced deep reinforcement learning (DRL) algorithms, and their integration into robotic and control systems. Beginning with the formalism of Markov Decision Processes (MDPs), the study outlines essential elements of the agent-environment interaction and explores core algorithmic strategies including actor-critic methods, value-based learning, and policy gradients. Emphasis is placed on modern DRL techniques such as DDPG, TD3, PPO, and SAC, which have shown promise in solving high-dimensional, continuous control tasks. A structured taxonomy is introduced to categorize RL applications across domains such as locomotion, manipulation, multi-agent coordination, and human-robot interaction, along with training methodologies and deployment readiness levels. The review synthesizes recent research efforts, highlighting technical trends, design patterns, and the growing maturity of RL in real-world robotics. Overall, this work aims to bridge theoretical advances with practical implementations, providing a consolidated perspective on the evolving role of RL in autonomous robotic systems.

Dynamic manipulation, such as robot tossing or throwing objects, has recently gained attention as a novel paradigm to speed up logistic operations. However, the focus has predominantly been on the object's landing location, irrespective of its final orientation. In this work, we present a method enabling a robot to accurately "throw-flip" objects to a desired landing pose (position and orientation). Conventionally, objects thrown by revolute robots suffer from parasitic rotation, resulting in highly restricted and uncontrollable landing poses. Our approach is based on two key design choices: first, leveraging the impulse-momentum principle, we design a family of throwing motions that effectively decouple the parasitic rotation, significantly expanding the feasible set of landing poses. Second, we combine a physics-based model of free flight with regression-based learning methods to account for unmodeled effects. Real robot experiments demonstrate that our framework can learn to throw-flip objects to a pose target within ($\pm$5 cm, $\pm$45 degrees) threshold in dozens of trials. Thanks to data assimilation, incorporating projectile dynamics reduces sample complexity by an average of 40% when throw-flipping to unseen poses compared to end-to-end learning methods. Additionally, we show that past knowledge on in-hand object spinning can be effectively reused, accelerating learning by 70% when throwing a new object with a Center of Mass (CoM) shift. A video summarizing the proposed method and the hardware experiments is available at https://youtu.be/txYc9b1oflU.

In this paper, we develop and open-source, for the first time, a square-root filter (SRF)-based visual-inertial navigation system (VINS), termed sqrtVINS, which is ultra-fast, numerically stable, and capable of dynamic initialization even under extreme conditions (i.e., extremely small time window). Despite recent advancements in VINS, resource constraints and numerical instability on embedded (robotic) systems with limited precision remain critical challenges. A square-root covariance-based filter offers a promising solution by providing numerical stability, efficient memory usage, and guaranteed positive semi-definiteness. However, canonical SRFs suffer from inefficiencies caused by disruptions in the triangular structure of the covariance matrix during updates. The proposed method significantly improves VINS efficiency with a novel Cholesky decomposition (LLT)-based SRF update, by fully exploiting the system structure to preserve the structure. Moreover, we design a fast, robust, dynamic initialization method, which first recovers the minimal states without triangulating 3D features and then efficiently performs iterative SRF update to refine the full states, enabling seamless VINS operation. The proposed LLT-based SRF is extensively verified through numerical studies, demonstrating superior numerical stability and achieving robust efficient performance on 32-bit single-precision floats, operating at twice the speed of state-of-the-art (SOTA) methods. Our initialization method, tested on both mobile workstations and Jetson Nano computers, achieving a high success rate of initialization even within a 100 ms window under minimal conditions. Finally, the proposed sqrtVINS is extensively validated across diverse scenarios, demonstrating strong efficiency, robustness, and reliability. The full open-source implementation is released to support future research and applications.

Chemistry, a long-standing discipline, has historically relied on manual and often time-consuming processes. While some automation exists, the field is now on the cusp of a significant evolution driven by the integration of robotics and artificial intelligence (AI), giving rise to the concept of the robochemist: a new paradigm where autonomous systems assist in designing, executing, and analyzing experiments. Robochemists integrate mobile manipulators, advanced perception, teleoperation, and data-driven protocols to execute experiments with greater adaptability, reproducibility, and safety. Rather than a fully automated replacement for human chemists, we envisioned the robochemist as a complementary partner that works collaboratively to enhance discovery, enabling a more efficient exploration of chemical space and accelerating innovation in pharmaceuticals, materials science, and sustainable manufacturing. This article traces the technologies, applications, and challenges that define this transformation, highlighting both the opportunities and the responsibilities that accompany the emergence of the robochemist. Ultimately, the future of chemistry is argued to lie in a symbiotic partnership where human intuition and expertise is amplified by robotic precision and AI-driven insight.

We present a deep reinforcement learning framework based on Soft Actor-Critic (SAC) for safe and precise maneuvering of double-Ackermann-steering mobile robots (DASMRs). Unlike holonomic or simpler non-holonomic robots such as differential-drive robots, DASMRs face strong kinematic constraints that make classical planners brittle in cluttered environments. Our framework leverages the Hindsight Experience Replay (HER) and the CrossQ overlay to encourage maneuvering efficiency while avoiding obstacles. Simulation results with a heavy four-wheel-steering rover show that the learned policy can robustly reach up to 97% of target positions while avoiding obstacles. Our framework does not rely on handcrafted trajectories or expert demonstrations.

The seamless integration of physical and digital environments in Cyber-Physical Systems(CPS), particularly within Industry 4.0, presents significant challenges stemming from system heterogeneity and complexity. Traditional approaches often rely on rigid, data-centric solutions like co-simulation frameworks or brittle point-to-point middleware bridges, which lack the semantic richness and flexibility required for intelligent, autonomous coordination. This report introduces the Knowledge Graph-Enhanced Multi-Agent Infrastructure(KG-MAS), as resolution in addressing such limitations. KG-MAS leverages a centralized Knowledge Graph (KG) as a dynamic, shared world model, providing a common semantic foundation for a Multi-Agent System(MAS). Autonomous agents, representing both physical and digital components, query this KG for decision-making and update it with real-time state information. The infrastructure features a model-driven architecture which facilitates the automatic generation of agents from semantic descriptions, thereby simplifying system extension and maintenance. By abstracting away underlying communication protocols and providing a unified, intelligent coordination mechanism, KG-MAS offers a robust, scalable, and flexible solution for coupling heterogeneous physical and digital robotic environments.

Camera-based 3D object detection and tracking are essential for perception in autonomous driving. Current state-of-the-art approaches often rely exclusively on either perspective-view (PV) or bird's-eye-view (BEV) features, limiting their ability to leverage both fine-grained object details and spatially structured scene representations. In this work, we propose DualViewDistill, a hybrid detection and tracking framework that incorporates both PV and BEV camera image features to leverage their complementary strengths. Our approach introduces BEV maps guided by foundation models, leveraging descriptive DINOv2 features that are distilled into BEV representations through a novel distillation process. By integrating PV features with BEV maps enriched with semantic and geometric features from DINOv2, our model leverages this hybrid representation via deformable aggregation to enhance 3D object detection and tracking. Extensive experiments on the nuScenes and Argoverse 2 benchmarks demonstrate that DualViewDistill achieves state-of-the-art performance. The results showcase the potential of foundation model BEV maps to enable more reliable perception for autonomous driving. We make the code and pre-trained models available at https://dualviewdistill.cs.uni-freiburg.de .

Successful generalist Vision-Language-Action (VLA) models rely on effective training across diverse robotic platforms with large-scale, cross-embodiment, heterogeneous datasets. To facilitate and leverage the heterogeneity in rich, diverse robotic data sources, we propose a novel Soft Prompt approach with minimally added parameters, by infusing prompt learning concepts into cross-embodiment robot learning and introducing separate sets of learnable embeddings for each distinct data source. These embeddings serve as embodiment-specific prompts, which in unity empower VLA models with effective exploitation of varying cross-embodiment features. Our new X-VLA, a neat flow-matching-based VLA architecture, relies exclusively on soft-prompted standard Transformer encoders, enjoying both scalability and simplicity. Evaluated across 6 simulations as well as 3 real-world robots, our 0.9B instantiation-X-VLA-0.9B simultaneously achieves SOTA performance over a sweep of benchmarks, demonstrating superior results on a wide axes of capabilities, from flexible dexterity to quick adaptation across embodiments, environments, and tasks. Website: https://thu-air-dream.github.io/X-VLA/

Efficient physics simulation has significantly accelerated research progress in robotics applications such as grasping and assembly. The advent of GPU-accelerated simulation frameworks like Isaac Sim has particularly empowered learning-based methods, enabling them to tackle increasingly complex tasks. The PAL Robotics TIAGo++ Omni is a versatile mobile manipulator equipped with a mecanum-wheeled base, allowing omnidirectional movement and a wide range of task capabilities. However, until now, no model of the robot has been available in Isaac Sim. In this paper, we introduce such a model, calibrated to approximate the behavior of the real robot, with a focus on its omnidirectional drive dynamics. We present two control models for the omnidirectional drive: a physically accurate model that replicates real-world wheel dynamics and a lightweight velocity-based model optimized for learning-based applications. With these models, we introduce a learning-based calibration approach to approximate the real robot's S-shaped velocity profile using minimal trajectory data recordings. This simulation should allow researchers to experiment with the robot and perform efficient learning-based control in diverse environments. We provide the integration publicly at https://github.com/AIS-Bonn/tiago_isaac.

This study investigates the developmental interaction between top-down (TD) and bottom-up (BU) visual attention in robotic learning. Our goal is to understand how structured, human-like attentional behavior emerges through the mutual adaptation of TD and BU mechanisms over time. To this end, we propose a novel attention model $A^3 RNN$ that integrates predictive TD signals and saliency-based BU cues through a bi-directional attention architecture. We evaluate our model in robotic manipulation tasks using imitation learning. Experimental results show that attention behaviors evolve throughout training, from saliency-driven exploration to prediction-driven direction. Initially, BU attention highlights visually salient regions, which guide TD processes, while as learning progresses, TD attention stabilizes and begins to reshape what is perceived as salient. This trajectory reflects principles from cognitive science and the free-energy framework, suggesting the importance of self-organizing attention through interaction between perception and internal prediction. Although not explicitly optimized for stability, our model exhibits more coherent and interpretable attention patterns than baselines, supporting the idea that developmental mechanisms contribute to robust attention formation.

Training robots to operate effectively in environments with uncertain states, such as ambiguous object properties or unpredictable interactions, remains a longstanding challenge in robotics. Imitation learning methods typically rely on successful examples and often neglect failure scenarios where uncertainty is most pronounced. To address this limitation, we propose the Uncertainty-driven Foresight Recurrent Neural Network (UF-RNN), a model that combines standard time-series prediction with an active "Foresight" module. This module performs internal simulations of multiple future trajectories and refines the hidden state to minimize predicted variance, enabling the model to selectively explore actions under high uncertainty. We evaluate UF-RNN on a door-opening task in both simulation and a real-robot setting, demonstrating that, despite the absence of explicit failure demonstrations, the model exhibits robust adaptation by leveraging self-induced chaotic dynamics in its latent space. When guided by the Foresight module, these chaotic properties stimulate exploratory behaviors precisely when the environment is ambiguous, yielding improved success rates compared to conventional stochastic RNN baselines. These findings suggest that integrating uncertainty-driven foresight into imitation learning pipelines can significantly enhance a robot's ability to handle unpredictable real-world conditions.

The true promise of humanoid robotics lies beyond single-agent autonomy: two or more humanoids must engage in physically grounded, socially meaningful whole-body interactions that echo the richness of human social interaction. However, single-humanoid methods suffer from the isolation issue, ignoring inter-agent dynamics and causing misaligned contacts, interpenetrations, and unrealistic motions. To address this, we present Harmanoid , a dual-humanoid motion imitation framework that transfers interacting human motions to two robots while preserving both kinematic fidelity and physical realism. Harmanoid comprises two key components: (i) contact-aware motion retargeting, which restores inter-body coordination by aligning SMPL contacts with robot vertices, and (ii) interaction-driven motion controller, which leverages interaction-specific rewards to enforce coordinated keypoints and physically plausible contacts. By explicitly modeling inter-agent contacts and interaction-aware dynamics, Harmanoid captures the coupled behaviors between humanoids that single-humanoid frameworks inherently overlook. Experiments demonstrate that Harmanoid significantly improves interactive motion imitation, surpassing existing single-humanoid frameworks that largely fail in such scenarios.

Embodied agents face a fundamental limitation: once deployed in real-world environments, they cannot easily acquire new knowledge to improve task performance. In this paper, we propose Dejavu, a general post-deployment learning framework that augments a frozen Vision-Language-Action (VLA) policy with retrieved execution memories through an Experience Feedback Network (EFN). EFN identifies contextually relevant prior action experiences and conditions action prediction on the retrieved guidance. We train EFN with reinforcement learning and semantic similarity rewards, encouraging the predicted actions to align with past behaviors under the current observation. During deployment, EFN continually expands its memory with new trajectories, enabling the agent to exhibit “learning from experience.” Experiments across diverse embodied tasks show that EFN improves adaptability, robustness, and success rates over frozen baselines. Our Project Page is https://dejavu2025.github.io/.

The dominant paradigm for training Large Vision-Language Models (LVLMs) in navigation relies on imitating expert trajectories. This approach reduces the complex navigation task to a sequence-to-sequence replication of a single correct path, fundamentally limiting the agent's ability to explore and generalize. In this work, we argue for and introduce a new paradigm: a shift from Path Imitation to Decision Understanding. The goal of this paradigm is to build agents that do not just follow, but truly understand how to navigate. We materialize this through two core contributions: first, we introduce Compass-Data-22k, a novel 22k-trajectory dataset. Its Reinforcement Fine-Tuning (RFT) subset provides a panoramic view of the decision landscape by annotating all feasible actions with A* geodesic distances. Second, we design a novel gap-aware hybrid reward function that dynamically adapts its feedback to decision certainty, shifting between decisive signals for optimal actions and nuanced scores to encourage exploration. Integrated into an SFT-then-RFT recipe, our CompassNav agent is trained not to memorize static routes, but to develop an internal compass that constantly intuits the direction to the goal by evaluating the relative quality of all possible moves. This approach enables our 7B agent to set a new state-of-the-art on Goal navigation benchmarks, outperforming even larger proprietary models, and achieve robust real-world goal navigation on a physical robot.

Generalist robot policies can now perform a wide range of manipulation skills, but evaluating and improving their ability with unfamiliar objects and instructions remains a significant challenge. Rigorous evaluation requires a large number of real-world rollouts, while systematic improvement demands additional corrective data with expert labels. Both of these processes are slow, costly, and difficult to scale. World models offer a promising, scalable alternative by enabling policies to rollout within imagination space. However, a key challenge is building a controllable world model that can handle multi-step interactions with generalist robot policies. This requires a world model compatible with modern generalist policies by supporting multi-view prediction, fine-grained action control, and consistent long-horizon interactions, which is not achieved by previous works. In this paper, we make a step forward by introducing a controllable multi-view world model that can be used to evaluate and improve the instruction-following ability of generalist robot policies. Our model maintains long-horizon consistency with a pose-conditioned memory retrieval mechanism and achieves precise action control through frame-level action conditioning. Trained on the DROID dataset (95k trajectories, 564 scenes), our model generates spatially and temporally consistent trajectories under novel scenarios and new camera placements for over 20 seconds. We show that our method can accurately rank policy performance without real-world robot rollouts. Moreover, by synthesizing successful trajectories in imagination and using them for supervised fine-tuning, our approach can improve policy success by 44.7%.

Current evaluation methods for autonomous driving prediction models rely heavily on simplistic metrics such as Average Displacement Error (ADE) and Final Displacement Error (FDE). While these metrics offer basic performance assessments, they fail to capture the nuanced behavior of prediction modules under complex, interactive, and safety-critical driving scenarios. For instance, existing benchmarks do not distinguish the influence of nearby versus distant agents, nor systematically test model robustness across varying multi-agent interactions. This paper addresses this critical gap by proposing a novel testing framework that evaluates prediction performance under diverse scene structures, saying, map context, agent density and spatial distribution. Through extensive empirical analysis, we quantify the differential impact of agent proximity on target trajectory prediction and identify scenario-specific failure cases that are not exposed by traditional metrics. Our findings highlight key vulnerabilities in current state-of-the-art prediction models and demonstrate the importance of scenario-aware evaluation. The proposed framework lays the groundwork for rigorous, safety-driven prediction validation, contributing significantly to the identification of failure-prone corner cases and the development of robust, certifiable prediction systems for autonomous vehicles.

Recent advancements in lunar positioning, navigation, and timing (PNT) have demonstrated that terrestrial GNSS signals, including weak sidelobe transmissions, can be exploited for lunar spacecraft positioning and timing. While GNSS-based navigation at the Moon has been validated recently, unmodeled ionospheric and plasmaspheric delays remain a significant error source, particularly given the unique signal geometry and extended propagation paths. This paper characterizes these delays using the Global Core Plasma Model (GCPM) and a custom low-cost ray-tracing algorithm that iteratively solves for bent signal paths. We simulate first-, second-, and third-order group delays, as well as excess path length from ray bending, for GNSS signals received at both lunar orbit and the lunar south pole under varying solar and geomagnetic conditions. Results show that mean group delays are typically on the order of 1 m, but can exceed 100 m for low-altitude ray paths during high solar activity, while bending delays are generally smaller but non-negligible for low-altitude ray paths. We also quantify the influence of signal frequency, geomagnetic $K_p$ index, and solar R12 index. These findings inform the design of robust positioning and timing algorithms that utilize terrestrial GNSS signals.