Papers for

search and rescue teams

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Snake robot moves and pushes objects using its whole body

Autonomous Whole-body Loco-manipulation Using a Snake Robot

Abstract: Snake robots can leverage their hyper-redundant morphology for both locomotion and manipulation through distributed body contact but whole-body loco-manipulation is particularly challenging since the entire body serves simultaneously as propulsive mechanism and manipulation tool Prior work on wheeled legged and snake robot platforms has decoupled these functions through dedicated appendages body segmentation or grippers We present closed-loop loco-manipulation with our snake robot COBRA without this distinction such that any part of the body may make contact We develop a contact-implicit dynamic model unifying robot-ground robot-object and object-ground interactions within a Moreau-Jean time-stepping framework and formulate a trajectory optimization over the full-order dynamics guided by a geometric planner and nominal proportional controller to autonomously push an object to a goal location during sidewinding We validate the planner and controller on hardware across varied configurations and present simulation results for a simplified trajectory optimization that resolves robot-object and object-ground contacts

Fri 6 NovRobotics
The gist
Moving and manipulating objects at the same time is hard for snake-like robots because they use their whole body for both tasks. The authors built a snake robot named COBRA that can touch and push objects anywhere along its body while moving. They created a detailed model and planning method to control the robot's movements and interactions with the ground and objects. They tested their approach on a real robot and in simulations, showing it can push objects to a target place while slithering.
Open → 2611.09777v1

Quadrotors optimized to change shape for flying through narrow gaps

Task-Driven Closed-Loop Design Optimization of Passively Morphing Quadrotors for Contact-Mediated Narrow Gap Traversal

Abstract: Aerial robots operating in cluttered environments may encounter openings narrower than their nominal width. Existing passively morphing drones reduce their width upon traversal, but they typically involve human-made designs, which can struggle to find the right balance among folding ratio, flight efficiency, and control tracking performance. We introduce a closed-loop design method that jointly optimizes quadrotor geometry and the torque-deflection response of two passively folding arms driven by the success rate in simulation for contact-mediated gap traversal. An optimization algorithm samples candidate designs that are evaluated in a contact-based simulation of the traversal task, where the task success rate is used as a feedback signal to sample progressively better designs over time. The resulting prototype passively reduces its width from 31.3 to 14.1 cm and traverses a 15 cm wide opening at a 2 m/s reference speed. It also tolerates gap approach-angle errors of up to 30 degrees and folds through a 40 cm long confined passage. Compared to a rigid twin, the soft configuration adds only 2 mm of position RMSE and less than 1% electrical power during contact-free flight. These results show that task-driven closed-loop design optimization can produce passively morphing airframes that are compact under contact, tolerant to approach error, and effectively rigid during nominal flight.

Fri 6 NovRobotics
The gist
Drones often need to fly through spaces smaller than their usual size, but it's hard to design them to shrink only when needed without hurting flight. The authors developed a method that tries out different drone arm designs in computer simulations, checking how well they can shrink and pass through tight openings without losing control. Their final drone can shrink its width by more than half to fit through a 15 cm gap while flying fast, and it still flies well when not squeezing through tight spaces. This approach helps make drones that automatically adapt their shape to tricky environments.
Open → 2611.09272v1

Millimeter wave sensing limits human pose accuracy for robots

What Limits mmWave Human-State Sensing for Robot Perception?

Abstract: Assistive robots and search-and-rescue platforms require reliable human state sensing in smoke, darkness, and privacy-sensitive environments, where cameras and LiDAR may be unreliable Millimeter-wave radar offers an alternative that works without ambient light, penetrates smoke, and does not capture visual appearance. However, the compact arrays that fit robot size, power, and cost budgets have limited apertures, producing joint-localization errors on the order of decimeters at a few meters, large enough to compromise the safety margins and tracking accuracy a robot depends on. Existing pipelines reduce pose error through frame fusion, multiple viewpoints, return selection, or stronger estimators, but examine one technique at a time, without relating its effect to a shared measurement limit or separating gains from added information from gains from estimator completion. We address this gap with a framework that models the human and environment as one time-varying scattering field and states a Cramer-Rao type error floor for joint localization as a function of bandwidth, aperture, signal-to-noise ratio, fused frame count, phase coherence, and visibility of retained returns. We then use existing frame fusion, multi viewpoint, return selection, and pose estimation methods as matched probes of this floor rather than as competing pipelines. Under the evaluated conditions, experiments on three datasets and a live deployment show these strategies move joint localization error in the predicted directions, for example lowering mean per-joint position error from about 97 to 65 mm with a second viewpoint, while a stronger estimator mainly lowers overall error by completing joints the measurement cannot resolve. These results give the robotics community a principled basis for deciding whether to prioritize observation design or estimator capacity when deploying reliable mmWave human sensing.

Fri 6 NovRoboticsHuman-Computer Interaction
The gist
When robots need to track human body positions in bad conditions like smoke or darkness, millimeter-wave radar can see through obstacles and protect privacy better than cameras. But small radar setups create position errors big enough to affect robot safety and tracking. The authors studied what causes these errors and how techniques like using multiple viewpoints or better calculations help. They found these methods help reduce errors somewhat, but there are fundamental limits based on the radar design and environment. Their work helps robot builders decide whether to improve the radar hardware or the data processing software for better human sensing.
Open → 2611.08863v1

Magnetic inertial system improves robot location where gps and cameras fail

A Tightly Coupled Magnetic-Inertial SLAM System for Field Robots

Abstract: Field robots operating in mines, tunnels, and disaster areas require reliable localization despite GNSS denial and degraded camera and LiDAR measurements caused by dust, smoke, or poor illumination. Missions involving both ground and aerial robots further motivate complementary sensing modalities that do not depend on visibility, preinstalled infrastructure, or wheel contact. This letter presents MISLAM, a tightly coupled magnetic-inertial simultaneous localization and mapping system using an IMU and a nine-magnetometer array, without relying on external odometry or platform-specific motion assumptions. The system combines the inertial navigation model with local polynomial magnetic-field models for short-term motion estimation and a global Gaussian process model for drift correction using an error-state Kalman filter. Offline evaluation on three indoor datasets collected with a handheld rig shows reductions in total endpoint closure error of 45.7% - 78.0% relative to magnetic-inertial odometry. Under poor illumination, MISLAM maintains tracking after visual-inertial odometry fails, achieving a horizontal closure error of 0.332 m, although vertical drift remains a limitation. These results support magnetic-inertial SLAM as a complementary localization modality for robotic sensor suites operating under degraded visual conditions.

Mon 2 NovRobotics
The gist
Robots working underground or in smoky, dark places often get lost because GPS and cameras don't work well there. The paper introduces a new system called MISLAM that uses sensors detecting magnetic fields and motion to help robots figure out where they are. The authors tested their system indoors and showed it can track location more accurately than older methods, especially when cameras can't see well. However, they noted vertical positioning still needs improvement.
Open → 2611.03106v1

OmniQuad robot moves easily with wheels and legs combined

OmniQuad: a Wheeled-Legged Hybrid Robot with Omnidirectional Wheels

Abstract: Combining wheeled and legged locomotion within a single robotic platform offers the potential to harness the advantages of both systems. The development of such hybrid systems remains an active area of research. While legged robots exhibit remarkable adaptability for traversing uneven and rugged terrains, their relatively low speed limits their practicality in indoor environments. Conversely, wheeled robots demonstrate higher efficiency on flat surfaces but often encounter difficulties when navigating obstacles such as steps. This paper introduces OmniQuad, a novel hybrid robot that integrates mecanum wheels with legged locomotion to exploit the benefits of both modalities. Mecanum wheels provide omnidirectional mobility, enhancing the robot's manoeuvrability by enabling agile movement in confined spaces and the ability to maintain a consistent orientation during task execution. In this work, we first present the design of a custom wheeled-foot featuring mecanum wheels. Subsequently, we describe the complete robot assembly, comprising these wheeled feet and EM-Act actuators. Finally, we detail the control scheme developed to govern the OmniQuad and evaluate its performance through experimental trials conducted in both indoor and outdoor environments.

Wed 28 OctRobotics
The gist
Robots that walk on legs are good at handling rough ground but move slowly indoors, while wheeled robots are fast on smooth floors but struggle with obstacles. The authors created OmniQuad, a robot that combines legs with special wheels that let it move in any direction. This design helps the robot move quickly inside buildings and also handle steps and uneven terrain outside. They built the robot, designed its controls, and tested it in different environments.
Open → 2610.38502v1

RAVEN system improves drone exploration under uncertain location tracking

RAVEN: A Unified Active SLAM System for Rapid Autonomous Visual Exploration with Aerial Robots under Localization Uncertainty

Abstract: Active simultaneous localization and mapping (SLAM) addresses the problem of planning robot motion to build complete and accurate maps of unknown environments through efficient exploration while maintaining high localization accuracy. This raises a dilemma between exploring unknown spaces to expand coverage and exploiting visited regions to perform active loop closures. Despite numerous prior works on ground robots, aerial robots with high agility and constrained sensor suites face more significant localization drift and thereby demand dedicated active SLAM planning. Moreover, a unified strategy that adapts performance based on task-specific priorities remains an open question. In this article, we introduce RAVEN, a unified active SLAM framework for Rapid Autonomous Visual ExploratioN using quadrotors, which systematically accounts for localization uncertainty. The framework starts with identifying frontiers and active loop closure candidates, from which global tour candidates are generated. These candidates are evaluated using a utility vector that separately quantifies exploration efficiency and localization accuracy, and the final tour is selected according to a prioritization parameter ranging from $[0,1]$. Finally, a perception-aware navigation planner generates smooth execution trajectories that minimize visual state estimation error based on the online Fisher information field that provides visual localization information in real-time. Benchmark evaluations demonstrate that RAVEN outperforms state-of-the-art active SLAM methods, highlighting its advantages in producing a spectrum of performance across prioritizations. Fully onboard field experiments with a customized stereo-inertial quadrotor further validate its practical applicability in challenging environments. The source code will be released to benefit the community.

Tue 27 OctRobotics
The gist
Mapping unknown places with flying robots is hard because they lose track of where they are, which hurts mapping quality. The authors created RAVEN, a system that helps drones explore more efficiently by balancing the search of new areas and revisiting known spots to correct location errors. It chooses routes by weighing exploration benefits against localization confidence and plans smooth paths that reduce visual errors during flight. Tests show this method improves mapping and location tracking compared to existing approaches, even on actual drones operating independently in challenging environments.
Open → 2610.37111v1

Humanoid robots improve parkour skills with better future step prediction

HC2Robot: Hyper-Connection Transformers with Predicted Foresight for Multi-Terrain Humanoid Parkour

Abstract: Humanoid parkour requires robots to reason about future footholds and whole-body motions over highly discontinuous terrain. We revisit the information asymmetry in asymmetric actor-critic learning for perceptive humanoid locomotion: the critic exploits privileged simulator signals during training, whereas the deployable actor operates solely from onboard perception. We argue that the key is not exposing privileged information to the actor, but learning a deployable approximation of it to narrow the information gap between training and deployment. We present HC2Robot, a deployable perceptive humanoid parkour framework consisting of a Predicted Foresight Module (PFM) and a Hyper-Connection (HC) Transformer. PFM predicts from deployable observations the policy-conditioned timing, location, and support quality of each foot's next touchdown event, converting privileged supervision into a deployable latent representation that provides the actor with predicted touchdown information without privileged runtime sensing. The HC Transformer further enhances temporal reasoning through learnable multi-stream routing, while a future motion prediction head provides short-horizon motion priors for regularization. Across stairs, gaps, high platforms, and slopes, HC2Robot achieves a mean simulated traversal success rate of 97.60% and consistently higher safe foothold rates than strong perceptive locomotion baselines, supporting improved traversal performance and geometric foothold support under deployable sensing.

Tue 27 OctRobotics
The gist
Parkour for robots means moving across tricky terrain like stairs and gaps without falling. The authors created a system called HC2Robot that helps a humanoid robot predict where and when it should place its feet next, using only the sensors a robot can carry with it. They use special training techniques that teach the robot to guess these important future steps even though it doesn’t have extra information during actual runs. This helps robots move more safely and successfully over difficult obstacles.
Open → 2610.36825v1

Humanoid robot adapts height to crawl and walk in tight spaces

REACH: An RL-based Hierarchical Humanoid Crawling Framework for Adaptable Confined Space Traversal

Abstract: Humanoid locomotion in confined spaces requires continuous height adaptation across multiple locomotion modes, yet most existing approaches remain limited to height-adjustable bipedal walking. We present REACH, an RL-based hierarchical framework that enables a Unitree G1 humanoid to traverse height-constrained environments through variable-height bipedal walking and crawling. The learned policies regulate body height over complementary command ranges of 1.27--0.50m in bipedal locomotion and 0.50--0.20m in crawling, extending traversal beyond the clearance limits of crouched walking. To address contact-rich crawling dynamics, we employ a history-conditioned latent representation that captures terrain and interaction context from proprioceptive observations. A bidirectional transition policy further connects crouched bipedal locomotion and crawling across large kinematic and contact discontinuities. These skills are integrated with a LiDAR-based planner that converts point clouds into collision-free three-dimensional paths and locomotion commands, including proactive height targets. Real-world experiments demonstrate accurate height and velocity tracking, robust terrain traversal, reliable mode transitions and confined-space traversal with the proposed pipeline. Code and videos are available at https://aislab-sustech.github.io/REACH/.

Tue 27 OctRoboticsArtificial Intelligence
The gist
Getting humanoid robots to move through very narrow spaces is hard because they need to change their height a lot and use different movement styles. The authors created a robotic system called REACH that teaches a Unitree G1 humanoid robot to both walk and crawl by changing its body height. This system uses smart learning methods to understand the robot’s environment and switch between walking and crawling smoothly. The robot can use laser sensors to plan safe paths and adjust its height to fit into small spaces. They tested this system in real life, showing the robot moves accurately and handles tricky terrain well.
Open → 2610.36546v1

Twisting sidewinding reduces motor load and speeds snake robots

Linear Twisting Sidewinding Gaits for Twistable Snake Robot Motion Planning on Complex Terrains

Abstract: Sidewinding is an efficient gait for snake robots to locomote on various terrains, by undulating the backbone curve and shifting the contact segments along it. However, relying solely on undulatory motion limits the locomotive capability of snake robots, and combining it with twisting motion can improve their speed and robustness. This hypothesis has recently been verified by the rolling vector approach, which locomotes the robot like a differential-drive car. Yet its twisting is concentrated in a short buffer segment, imposing most of the workload on the two twist joints in the middle, which can often overburden the motors and limit its locomotion. With that in mind, this paper develops a new family of twisting sidewinding gaits that distribute the workload over the entire body with our proposed linear twisting function. Under matched turning speed and total output power, our gaits reduce the peak output power of a single twist joint by 42%. Real-robot experiments show that, with the reduced workload, our new gaits can locomote the robot in various terrains that the existing rolling vector approach cannot. Besides, our new gaits can move up to 49% faster than conventional sidewinding without twisting.

Wed 21 OctRobotics
The gist
Snake robots move by mimicking real snakes using a sidewinding gait, but this can strain some motors and slow the robot down. The authors propose a new way to add twisting motion evenly along the robot’s body, rather than concentrating it in a few joints. This reduces the motor effort needed by 42%, lets the robot move faster by up to 49%, and allows it to travel on rough terrain where older methods failed. Their experiments tested these new movements on actual snake robots.
Open → 2610.29461v1

Cable tension sensing improves safety of tethered robot movement

Tension Tells the Tail: Cable-Obstacle Interaction Sensing for Tethered Mobile Robot Navigation

Abstract: Tethered mobile robots can sustain long-duration operation through continuous power supply and stable communication via their cables. However, cable-environment interaction causes unintended obstacle displacement, cable snagging or entanglement, and even cable damage, leading to mission failure. To assess the risk of individual cable-obstacle contacts, this study investigates cable tension as a sensing modality for estimating whether a contacted obstacle is movable or non-movable. We conducted experiments with a tethered mobile robot and collected tension data while varying robot velocity, contact angle between the obstacle and cable, obstacle mass, and obstacle side-surface material. The results show that obstacle movability is characterized by the initial tension rise and, more importantly, by whether the tension relaxes or remains high after its peak. Based on these observations, we select six features from the tension profiles and train a logistic-regression classifier using 117 labeled trials. The proposed method distinguishes movable and non-movable cases with a balanced accuracy of 95.5% after 30 cm of post-contact travel. These results demonstrate the potential of cable tension for risk assessment of cable-obstacle interactions and provide a step toward online, cable-aware autonomous navigation.

Sat 17 OctRobotics
The gist
Tethered robots use cables for power and communication, but these cables can get caught or damage obstacles during movement. The authors show that by measuring how the cable pulls or relaxes when touching an obstacle, a robot can tell if the obstacle can be moved or not. They trained a computer model on tension data from real robot experiments and achieved over 95% accuracy in identifying obstacle movability. This helps robots avoid problems like snagging and improves their ability to navigate safely.
Open → 2610.24449v1

Stabilizing flipper control helps tracked robots avoid getting stuck without terrain maps

Mapless, Stabilizing, and Anti-Stuck Flipper Control for Articulated Tracked Robots

Abstract: Nearly every autonomous or semi-autonomous flipper controller assumes access to an elevation map of the terrain under and in front of the robot. The terrain beneath the chassis is occluded in the sensors, so the map is remembered through odometry integration, whose drift corrupts it. We present Stabilizing Anti-Stuck (SAS), a mapless controller that combines a stabilization term for body roll and pitch with an anti-stuck term that frees the robot when it gets stuck. As input, SAS uses the IMU, the flipper joint angles, and a motion cue that only tells whether the robot is moving. The cue is computed over a short window and never accumulated, so SAS does not inherit the odometry-integration drift that corrupts the map. We deploy SAS on a tracked robot with four flippers and evaluate it against manual teleoperation and prior flipper controllers in a large comparison, both in simulation and on a real obstacle. In simulation and on the real robot, SAS achieves traversal quality comparable to state-of-the-art map-based controllers. On a log hidden in tall grass, invisible to the LiDAR, the best map-based controller gets stuck while SAS traverses the hidden obstacle without a problem.

Fri 9 OctRobotics
The gist
Tracked robots use flippers to move over tricky obstacles, but they usually rely on maps of the ground to control these flippers safely. The authors developed a new method called SAS that doesn’t need maps at all. Instead, it uses sensors to keep the robot balanced and to detect when it’s stuck, helping it get free. Their tests showed SAS works as well as map-based methods and even succeeds where maps fail, like when obstacles are hidden in tall grass.
Open → 2610.13838v1

Multi-modal UAV system improves maritime vessel tracking accuracy

Multi-Modal Graph Neural Network-Inspired Vessel Identification and Tracking for UAV-Based Maritime Surveillance

Abstract: The unmanned aerial vehicle (UAV)-based maritime surveillance is becoming an important part of maritime Internet of Things (IoT) and intelligent transportation systems. It provides a flexible and cost-effective way to large-scale maritime traffic monitoring. However, achieving robust situational awareness remains challenging in practical maritime environments. The main challenges arise from heterogeneous sensing modalities, adverse imaging conditions, and spatial misalignment caused by UAV motion. To address these problems, this paper proposes an airborne multi-modal graph learning-driven framework, termed AMGT, for vessel identification and tracking in UAV-assisted maritime IoT. Specifically, visible and infrared images are first aligned and fused through a multi-parameter perspective transformation model. This improves visual perception under low visibility and complex illumination conditions. Automatic identification system (AIS) messages are then temporally synchronized using interpolation and linear prediction. To detect small and oriented vessels, we develop a mixed-modality MAir-YOLOv8-OBB detector, followed by the MAir-SORT algorithm for stable vessel tracking. We further design a graph learning-driven association network to model temporal dynamics and spatial dependencies among vessels and between vessels and UAVs. A motion-aware coordinate unification (MCU) strategy is also introduced to reduce nonlinear misalignment caused by UAV motion. Experiments on the MAP-M2UAV50 dataset show that AMGT achieves high accuracy and efficiency. The results demonstrate its potential for UAV-assisted maritime IoT applications, including vessel monitoring, traffic situational awareness, and intelligent maritime transportation services. The dataset and code will be available at https://github.com/MAPGroup/AMGT.

Fri 9 OctComputer Vision and Pattern Recognition
The gist
Monitoring ships from drones is useful but hard because of bad weather, different cameras, and drone movement. The authors created a new method that combines visible and infrared images and synchronizes ship data for better detection and tracking. Their approach uses smart computer models that understand ship movements and positions over time, making it easier to follow small or hard-to-see vessels. Tests show this method works well for real-life maritime monitoring.
Open → 2610.13220v1

Humanoid robots navigate without vision using touch and planning

TAPNAV: Humanoid Navigation through Tactile Active Perception

Abstract: Navigation in vision-denied environments is challenging for humanoid robots because proprioceptive odometry drifts and localization uncertainty accumulates rapidly. We present TAPNAV, a tactile active-perception framework that enables humanoid navigation toward a goal by actively probing surrounding structures without relying on vision. TAPNAV maintains a pose belief from odometry, IMU, and tactile contact observations, and couples uncertainty-aware global route planning with information-gain-driven local probing. The global planner searches for routes that keep predicted localization uncertainty bounded by exploiting opportunities for tactile correction, while the local planner selects probe actions that maximize expected information gain. A whole-body controller coordinates the humanoid's locomotion and end-effector contact to execute the planned navigation and probe motions. We evaluate TAPNAV in simulation and on a Unitree G1 across different floor plans and obstacle geometries. TAPNAV achieves lower state estimation error and a higher task completion rate than baselines. These results demonstrate that actively planning physical interactions with the environment can provide localization cues for reliable humanoid navigation without vision.

Wed 7 OctRobotics
The gist
Humanoid robots normally use cameras to find their way, but in places where vision doesn't work, they get lost quickly. The authors developed a system called TAPNAV that helps robots move toward their goal by feeling their surroundings with touch instead of relying on sight. TAPNAV combines information from the robot's movement sensors and touch to keep track of its position and plans routes and probing moves that reduce uncertainty. The system was tested in simulations and on a real robot, where it outperformed other methods by completing tasks more reliably.
Open → 2610.10748v1

Robots map and inspect underground mineral deposits for mining

Towards Spatial Perception for Heterogeneous Robot Collaboration in Subterranean Mining Environments

Abstract: The autonomous extraction of deep mineral deposits in abandoned underground mines is fundamentally a multi-agent integration problem. No single platform simultaneously offers the mobility to traverse kilometers of degraded drifts and the sensing payload required to characterize an ore body. This article presents the onboard perception pipeline that bridges two heterogeneous agents within the PERSEPHONE autonomous mining mission. Which consist of a lightweight Explorer robot that maps an unknown mine and generates a 3D scene graph of inspection targets, by running a zero-shot, vision-language semantic segmentation stack that detects mineral deposits directly from natural-language prompts. The map and the graph are then handed to a second Inspector robot, which carries an advanced sensing payload and uses them to plan close-range inspection viewpoints. We detail the complete pipeline, with emphasis on the geometric abstraction that turns raw detections into actionable inspection targets, spanning per-view bounding-box generation, cross-view box merging, plane fitting, and polygon extraction, and we report an extensive field validation in a subterranean test facility and in an active magnesite mine, covering both iron-vein and magnesite mineralization under realistic, perceptually degraded conditions.

Tue 29 SeptRobotics
The gist
Exploring deep abandoned mines is tough because no single robot can both explore long tunnels and analyze minerals well. The authors describe two robots working together: one maps the cave and finds minerals using an AI that understands language, while the other inspects the minerals more closely using the map and data from the first robot. They tested this approach underground, showing it works even in difficult conditions like poor lighting and dust.
Open → 2609.37419v1

AerialDojo builds large benchmark for aerial object search tasks

AerialDojo-200K: A Large-Scale Benchmark Suite for Open-World Aerial Object-Goal Search

Abstract: Open-world aerial object-goal search is a foundational yet challenging task, requiring aerial agents to autonomously explore large-scale, unstructured three-dimensional environments and reach target objects specified by semantic descriptions or reference images, rather than following route-specific instructions. However, research in this task remains at a nascent stage and relies on small, environment-specific benchmarks with heterogeneous action spaces and data formats. These limitations hinder large-scale training and cross-benchmark evaluation, constraining the scalability and generalizability of aerial agents. To address this problem, we propose AerialDojo-200K, a large-scale benchmark suite for open-world aerial object-goal search, with 3 times as many scenes and 18.7 times as many task instances as the largest existing benchmark for this task. Specifically, we construct 42 simulation scenes spanning four scene families and 21 scene types, including 18 urban, 12 natural, six infrastructure, and six disaster scenes. To ensure data quality, 12 annotators spent two months manually annotating 109 landmarks, 2099 target objects, and 2099 object anchors across these scenes. We further construct 205,732 task instances, comprising over 100K semantic-goal and over 100K image-goal instances across Base, Standard, and Long-Horizon settings. Each task instance includes a collision-free reference trajectory and corresponding multi-view video recordings. We also develop a unified evaluation framework with a scene partition comprising 21 in-distribution scenes and 21 out-of-distribution scenes. Finally, our evaluation of five open-source and four closed-source multimodal large language models reveals that there is still a long way to go toward achieving general-purpose aerial agents. All can be found at https://fengtt42.github.io/AerialDojo/.

Mon 28 SeptComputer Vision and Pattern RecognitionArtificial IntelligenceEmerging Technologies
The gist
Finding objects from the air in large, open places is hard because drones must explore many kinds of environments without simple directions. The authors created AerialDojo-200K, a huge collection of simulated scenes and search tasks that covers cities, nature, and disaster areas to help train and test aerial search systems better. They carefully labeled thousands of landmarks and targets, and included video data and clear paths for each task. They also tested various advanced language-vision models and found current AI still struggles with this problem.
Open → 2609.36066v1

Bipedal robot controller for walking on varied and tricky surfaces

Multi-Terrain Mastery: A Comprehensive Controller for Bipedal Locomotion

Abstract: Advancing bipedal robots to navigate diverse terrains remains a significant challenge in robotics. Traditional locomotion controllers excel on specific surfaces but struggle across varied environments, limiting their practical applications. Given the unpredictable nature of real-world environments, a single controller capable of handling multiple terrains is ideal, eliminating the need for multiple specialized controllers. We propose a multi-terrain controller to enhance the versatility and robustness of bipedal locomotion. Building on previous work with a stance ankle motor for stability on inclined and rough surfaces, this paper extends capabilities to steep wet uneven grassy slopes, and compliant terrains such as sand, gravel, rocks, and constrained terrains like staircases. To address the unique demands of these terrains, we introduce a new impact map that is essential for maintaining performance and robustness against unseen terrains. We also discuss in detail the control structure for real-time deployment on the robot. We validate our controller on the 20 degree-of-freedom Cassie bipedal robot.

Sun 27 SeptRobotics
The gist
Walking robots usually handle only one kind of ground, making it hard for them to move around in real places with many different surfaces. The authors created a single controller that helps a two-legged robot walk steadily on many surfaces, like wet grass, sand, rocks, and stairs. They improved the robot’s stability by adding a new way to handle impacts from walking on uneven or soft grounds. Their controller was tested on a complex robot called Cassie with many moving parts, showing it works in real time.
Open → 2609.33174v1

Robot climbs vertical surfaces by jumping with optimized shape and motion

Co-design of trajectory and morphology for a vertical jump-climbing robot

Abstract: Animals such as squirrels and even bears have adapted to rapidly climb up trees and other complex vertical terrain, but achieving comparable agility has been a challenge for climbing robots. Existing robots often use walking gaits and move conservatively to stay in contact with the surface, which limits the range of dynamic maneuvers. In this paper we present a 290 g robot that, to our knowledge, is the first to climb vertically by bounding (with an aerial phase). We leverage a co-design workflow, in which the morphology and trajectory are jointly optimized for fast locomotion, subject to adhesion force limitations seen in spined grippers. The resulting trajectory includes a rapid maneuver that launches the robot vertically, and an aerial reorientation that brings the front grippers back to the surface using the rear leg as an inertial tail. We evaluate the resulting jump forces in 2D force space and demonstrate that the optimized morphology is capable of continuous climbing at a speed of 0.375 m/s (1.97 body lengths/s), and can also achieve ground locomotion and transition to a vertical surface. Our work proposes design insights for the jump-climbing maneuver and serves as an important step toward creating climbing robots with agility on par with that of animals.

Fri 25 SeptRobotics
The gist
Climbing robots usually move slowly and carefully to stay stuck to walls, making it hard for them to jump or move quickly like squirrels or bears. The authors built a lightweight robot that can jump up vertical surfaces by combining its shape and how it moves for a smooth jump and landing. Their robot uses its back leg as a sort of tail to turn itself around mid-air and land safely with its front grippers. This design lets the robot climb continuously at nearly two body lengths per second and also walk or jump onto vertical walls.
Open → 2609.31014v1

Snake robots with more joints move better through obstacles

Dense-Joint-Based Obstacle-Aided Locomotion with a Joint-Repositionable Snake Robot

Abstract: Obstacle-aided locomotion is a fundamental capability for snake robots to traverse complex environments. However, conventional rigid-link snake robots often suffer from stagnation or jamming caused by their low joint density (i.e., the number of joints per unit length). This results in discontinuous contact with obstacles, unlike the continuous adaptation of biological snakes. To investigate the effect of joint density on obstacle-aided locomotion performance, we utilized a joint-repositionable snake robot mechanism that decouples actuators from joints, enabling a high-density architecture. We developed two experimental models with identical total lengths but different joint densities (high-density and low-density) and conducted comparative propulsion experiments in obstacle environments with varying obstacle diameters. The experimental results demonstrate that the high-density model substantially suppresses the abrupt shifts in reaction forces that cause stagnation in the low-density model. By maintaining smooth contact points, the high-density configuration reduces power consumption and achieves stable, continuous propulsion. These results highlight high joint density as a key factor in improving the environmental adaptability of snake robots in complex terrains.

Thu 24 SeptRobotics
The gist
Robots that move like snakes often get stuck when trying to squeeze through complicated spaces because they don’t have enough joints along their bodies. The researchers built a new kind of snake robot that can change where its joints are along its body, allowing it to have more joints packed closely together. They found that having many joints helps the robot stay in smooth contact with obstacles, so it moves more steadily and uses less power. This approach makes snake robots better at moving through tricky environments, like cluttered rooms or rough outdoor terrain.
Open → 2609.29261v1

Compact spherical robot rolls and propels in water with internal moving masses

Omnidirectional Amphibious Locomotion via Internal Mass Actuation

Abstract: Field robots must traverse varied terrain and obstacles while remaining robust to water, debris, vegetation, and physical contact. We present MARBLE, a fully enclosed omnidirectional amphibious rolling robot driven entirely by internal mass redistribution. Three mutually orthogonal linear sliders shift internal masses to generate body rotation, while an orientation-aware controller maps planar velocity commands into slider positions. A rigid spherical shell encloses all active mechanisms and simultaneously serves as the terrestrial contact surface, buoyant enclosure, and mounting structure for passive fins that enable water-surface propulsion. Rotation of the same shell architecture hence produces rolling on land and surface propulsion in water without mechanical reconfiguration or separate locomotion actuators. The spherical morphology further allows the robot to accommodate changes in body orientation and contact location during direct interactions with terrain and obstacles. We evaluate MARBLE through omnidirectional locomotion characterization, traversal across heterogeneous terrestrial environments, aquatic surface locomotion, land-water transitions, and deliberate obstacle interactions. These experiments demonstrate how a single enclosed mechanical architecture can combine omnidirectional mobility, cross-medium locomotion, and tolerance to environmental contact. MARBLE provides a compact design for field mobility across heterogeneous terrain, obstacles, and land-water transitions. We will open-source all software and hardware design. Our website is https://generalroboticslab.com/MARBLE

Wed 23 SeptRobotics
The gist
robots that work outdoors must move across land, water, and obstacles without breaking. The authors built MARBLE, a ball-shaped robot that moves by shifting weights inside itself instead of using wheels or propellers. Its shell rolls on land and also floats with fins to glide on water, all without changing parts. This design lets it move in any direction and handle rough environments and obstacles smoothly. The authors tested it on land, water, and transitions between both to show it works well.
Open → 2609.27358v1

Multi-robot exploration improves with probabilistic peer intent sharing

Decentralized Multi-Robot Exploration with Probabilistic Peer Intent and Multi-hop Plan Propagation

Abstract: Efficient coordination under limited communication remains a key challenge in decentralized multi-robot exploration. While centralized approaches benefit from global information sharing, they are often impractical in large-scale or communication-constrained environments. Existing Monte Carlo Tree Search (MCTS)-based approaches, such as Decentralized Monte Carlo Exploration (DMCE), enable decentralized planning by taking peer intent into account. This peer intent is obtained by communicating sequences of planned waypoints with robots within direct communication range. In this work, we extend this idea by introducing Probabilistic Peer Intent (PPI), which converts peer trajectories into a continuous spatial representation of predicted intent and incorporates it into local MCTS action evaluation. We additionally study the effects of sharing peer intent beyond direct communication range by propagating plans over multiple hops. Experiments across multiple simulated environments and team sizes show that PPI and Multi-hop propagation can each improve decentralized exploration, with their relative benefits depending on environment structure and team size. We also demonstrate the real-world deployment of our method on three robots operating in different environment types.

Sat 19 SeptRoboticsMultiagent Systems
The gist
Coordinating multiple robots to explore an area can be hard when they cannot communicate easily. The authors developed a way for robots to share their planned paths in a way that predicts where others might go, even beyond direct communication range. This helps each robot plan better and explore more efficiently without needing constant contact. They tested this approach in simulations and with real robots in different environments.
Open → 2609.22726v1

Safe multi-robot area coverage for cars with movement limits

Density-Driven Area Coverage for Nonholonomic Multi-Robot Systems with Safety Guarantee

Abstract: Density-Driven Optimal Control (D2OC) provides a principled approach to distributing multi-robot teams over non-uniform spatial distributions. Applying D2OC to nonholonomic robots, however, creates a gap between safety constraints imposed on a reference motion and the physical inputs that determine the actual robot motion. We address this issue by enforcing the safety constraint directly on the robot's physical inputs while preserving the density-driven coverage objective. The proposed framework combines D2OC with a control barrier function safety filter through a feedback-linearizing look-ahead point, allowing safety and actuator limits to be considered together during control. We further derive a safety margin that accounts for the look-ahead geometry, robot footprint, and motion during each control interval. Simulation results show that the proposed method maintains the required physical separation while achieving coverage performance comparable to a conventional reference-tracking approach, which can satisfy safety on the reference motion yet violate the corresponding physical clearance. Experiments on multiple nonholonomic robots in the Robotarium further demonstrate safe execution while driving the robots toward the desired spatial distribution. These results show that enforcing safety directly on the physical inputs can eliminate the mismatch between safety certification and physical robot motion in density-driven multi-robot coverage.

Sat 19 SeptRoboticsMultiagent Systems
The gist
Getting many robots to spread out safely over an area is tricky when the robots can't move directly sideways, like cars that must turn to change direction. The authors found that controlling safety rules on the actual driving commands, rather than on a simpler planned path, stops robots from getting too close while still covering the area efficiently. They combined a special safety filter with the robot’s controls and calculated safe distances considering the robot’s shape and how it moves. They tested this method in simulations and real robot experiments, showing it works well to keep robots safe while guiding them where they need to be.
Open → 2609.22668v1

Autonomous drone recovery system enables precise midair multirotor docking

Integrated Guidance and Control of a Mother-Child UAV-UGV System for Cooperative Missions

Abstract: Autonomous recovery of a small multirotor onto a hovering multirotor carrier differs from recovery onto ground or shipborne platforms because the recovery surface is itself an actively controlled, thrust-limited aerial vehicle. This paper presents a field-validated autonomy framework for a heterogeneous rover-mothership-child system executing rover supervision, mothership transit, child deployment and sortie, autonomous return, aerial recovery, and synchronized descent. The recovery stack combines jerk-bounded reference generation, disturbance-observer-augmented planar tracking, feasibility-aware vertical control, a discrete-time barrier-based safety filter for relative vertical geometry, and communication-aware carrier-state prediction. The contribution is the coordinated system-level integration of these methods for recovery onto a hovering multirotor and its full-scale outdoor validation. The framework is implemented on a PX4-ROS 2 architecture using RTK-enabled GNSS, IMU, and barometric fusion, with mothership-side 1D lidar used only as an auxiliary near-contact cue. RTK-fixed positioning was maintained throughout testing. Across 20 outdoor cooperative missions, 17 successfully completed deployment, sortie, and recovery, giving an observed mission success rate of 85%. For successful recoveries, mean terminal-alignment time was 6.3 s, mean planar alignment error at acceptance was 0.18 m, maximum terminal planar deviation was 0.32 m within a 0.40 m capture radius, and minimum logged relative vertical separation during coupled descent was 0.41 m. Mothership planar station-keeping RMS error was 0.25 m. The three unsuccessful trials occurred at different mission stages and are analyzed separately. Results demonstrate practical autonomous aerial recovery within the tested outdoor operating envelope.

Thu 17 SeptRobotics
The gist
Landing a small drone onto another hovering drone is tricky because the landing surface moves and uses its own thrust. The authors developed a system where a larger drone (the mothership) carries and launches a smaller drone (the child), which then returns and lands on the mothership autonomously. Their approach uses advanced control techniques and sensors to keep both drones aligned and safe during landing. They tested the system outdoors 20 times, achieving an 85% success rate in complete missions.
Open → 2609.20540v1

Vine robots grow safely by finding paths with minimum pressure

An Efficient Algorithm for Minimum-Pressure Growth Planning of Vine Robots

Abstract: Vine robots navigate cluttered environments by extending from their tip. Although their ability to operate in such environments has been extensively demonstrated, little work has addressed growth planning, i.e., finding optimal growth paths. Moreover, existing planners do not account for the growth pressure necessary to follow a given path, which can cause the robot to burst when it is too high. In this paper, we address the problem of finding minimum-pressure paths for vine robots growing around polytopic obstacles. We propose an efficient algorithm that is guaranteed to find globally optimal solutions in 2D and approximate solutions in 3D, with an error that vanishes as a discretization parameter approaches zero. First, we derive a growth pressure equation for vine robots of arbitrary shape, which we use to show that there always exists a minimum-pressure path that is piecewise-linear and can bend only at specific points on the obstacles. We then leverage this observation to reduce the growth-planning problem to a shortest-path problem with time-dependent weights, which we efficiently solve using a modified Dijkstra's algorithm. We demonstrate the speed and scalability of our approach through numerical simulations. We also validate our algorithm with hardware experiments and provide an open-source and high-performance implementation in the Python package, VinePlanner: https://github.com/Ahsoka/VinePlanner.

Wed 16 SeptRobotics
The gist
Vine robots are soft robots that move by growing at their tips, which makes navigating through cluttered spaces tricky because too much pressure can cause them to burst. The authors developed a new method to calculate the safest path for these robots by minimizing the pressure needed to grow along a route. Their approach finds the best paths in two dimensions and good approximations in three dimensions, and it works quickly by focusing on certain key points near obstacles. They tested their method both in computer simulations and real experiments, and they also shared their software tool so others can use it.
Open → 2609.18070v1

Hierarchical path planning improves robot coverage in unknown areas

A Hierarchical Coverage Path Planning Algorithm for Unknown Environments

Abstract: This paper presents an online coverage path planning algorithm for unknown environments. During navigation, the initially unknown search area is progressively decomposed into disconnected subareas as new obstacle information is acquired and coverage proceeds. These subareas are organized in an incrementally constructed decomposition tree that preserves their hierarchical parent-child relationships. Based on this tree, a global coverage tour is maintained and updated online by prioritizing newly generated child subareas according to their exploration states and distances from the robot. A local planner then generates coverage motions within each selected subarea, allowing the robot to adapt its trajectory as the environment is gradually revealed. Its performance is evaluated via high-fidelity simulations in complex scenarios. The results show improved coverage efficiency in terms of path length and overlap ratio in comparison to three baseline algorithms.

Fri 11 SeptRobotics
The gist
Robots often need to explore and cover areas they haven't seen before, like rooms filled with obstacles. This paper introduces a way for robots to break down these unknown spaces into smaller parts while they move around. The robot keeps track of these parts in a tree structure to decide the best order to explore them. This helps the robot cover the area more efficiently by avoiding unnecessary backtracking or overlaps, as shown in computer simulations.
Open → 2609.12595v1

Multi-robot teams improve exploration by smart communication timing

Communication-Constrained Multi-Robot Exploration With Adaptive Communication Windows

Abstract: Exploring unknown environments with multi-robot teams can improve efficiency by allowing robots to explore in parallel. However, realizing these gains requires effective information sharing. When communication is intermittent, robots must balance the benefits of sharing information against the cost of diverting from exploration to establish communication. This paper introduces MACE, a decentralized exploration framework that actively evaluates whether establishing communication is worthwhile. At scheduled communication windows, robots estimate the cost of reaching previously identified communication locations. By formulating this decision as a variant of the Vehicle Orienteering Problem, robots evaluate routes based on the travel required to establish communication and the exploration that can be completed along the way. This approach enables robots to communicate more frequently than under purely opportunistic strategies while reducing the unnecessary travel associated with fixed rendezvous strategies. Across a set of simulated environments with varying size and geometry, we demonstrate that MACE reduces the total exploration time by up to 23% compared to existing communication-constrained exploration strategies.

Fri 11 SeptRobotics
The gist
Exploring unknown areas with many robots can be faster, but robots need to share what they find. Sharing information takes time and effort, especially when communication isn’t always available. The authors created a system called MACE that helps robots decide the best times to connect and share without wasting time traveling just to meet. This approach made exploration up to 23% faster in computer simulations compared to older methods.
Open → 2609.12502v1

Humanoid robot learns to swing across bars like primates

SwingBot: Learning Whole-Body Brachiation for Humanoid Robots

Abstract: Brachiation enables primates to move across overhead supports when ground paths are blocked, suggesting a complementary locomotion mode for robots operating in cluttered or hazardous environments. Bringing this capabil?ity to high-DoF humanoid robots is difficult because the controller must discover a long-horizon release-swing-capture sequence, coordinate alternating contacts with whole-body momentum, and act without reliable measurements of segment?relative displacement or hook-contact state. We present SwingBot, a learning framework for continuous humanoid brachiation with passive wrist hooks. Swing?Bot makes the task trainable by organizing learning around the structure of brachi?ation: biomimetic keyframes make rare release-swing-capture transitions reach?able during early exploration, and recurrent privileged-state estimation provides compact position and contact latents for deployment. Hardware experiments demonstrate continuous bar traversal and robustness to payload, external distur?bances and different bar spacings, showing that this formulation offers a practical route to whole-body robotic brachiation.

Wed 9 SeptRobotics
The gist
Moving through cluttered spaces can be hard for robots that walk on two legs. The authors developed SwingBot, a system that helps humanoid robots swing hand over hand, like monkeys on jungle gyms. Their approach breaks down this complex swinging motion into simpler steps and uses special learning to train the robot to keep its balance and grip. They tested SwingBot on real robots swinging through bars repeatedly, even with added weight and bumps, showing it can handle tricky movements.
Open → 2609.10283v1