HC2Robot proposes a deployable perceptive humanoid parkour framework composed of a Predicted Foresight Module (PFM) and a Hyper-Connection (HC) Transformer. The PFM predicts, from deployable observations, the timing, location, and support quality of each foot's next touchdown event, converting privileged supervision into a deployable latent representation.
For the high-school remote-control division of the 2027 NASA Human Exploration Rover Challenge, a study presents an integrated, reproducible design assessment of a four-wheel educational rover: 10.919 kg, 215 mm wheels at 200 RPM give a theoretical rolling speed of 8.105 km/h, and the nominal motor specifications are insufficient to demonstrate compliance with the 8 km/h design ceiling.
An arXiv paper proposes modeling bio-inspired magnetic navigation (BIMN) as a gradient flow control and guidance problem, using magnetic inclination and intensity of Earth's magnetic field to guide a system through intermediate waypoints to a destination within a predefined region. The authors note that magnetic coordinates do not map one-to-one to geographic locations and are ambiguous, but that gradient flow can act as a "water current" carrying the system's motion and be steered. The framework could be used in engineering to build geomagnetic navigation infrastructure and to define and follow prescribed routes, and also to analyze animal migration routes that rely on magnetoreception.
The authors propose QuadPose, which estimates the full-body 18-DoF pose of a quadruped robot from monocular video sequences, including the 6-DoF base pose and 12-DoF joint angles, and release QuadSyn, the first camera-to-quadruped robot pose benchmark. On QuadSyn, QuadPose achieves 92.18% PCK@2.5px and 85.94% ADD@20cm, with a mean joint-angle error of 3.65 degrees and a runtime of about 33 FPS.
OpenJaws is a fully open-source sensorized UMI hardware and software framework that pairs a modular 3D-printed gripper with a low-cost stereo camera to recover metric camera trajectories, improving action-label localization accuracy in handheld demonstrations. The authors also propose UMI-Loc-Bench-v1, a benchmark for quantifying the real-world localization accuracy of handheld data-collection devices, and show that stereo localization clearly outperforms monocular alternatives. Policies trained on data collected with OpenJaws have been deployed on real robots across multiple tasks, and the hardware, software, and data are released under a commercially friendly license.
GRAFT proposes a geometry-aware residual adaptation framework that, without changing the parameters of a frozen humanoid behavior foundation model (BFM) or its motion-reference interface, uses a lightweight adapter to apply terrain-conditioned corrections only to lower-body execution targets. The adapter is trained with task and foot-geometry rewards, is regularized so it produces no unnecessary corrections on flat ground, and is distilled from privileged terrain observations to onboard depth.
SVIn+ is a multi-sensor state estimation framework for autonomous underwater vehicles (AUVs). Building on the open-source visual-inertial SLAM system SVIn, it integrates water-pressure and DVL measurements into tightly coupled nonlinear optimization and uses a KISS-ICP decoupled front end to generate relative pose factors, incorporating 3D sonar structural information into the factor graph. The authors conducted field deployments in shipwreck and coral reef scenarios, validating drift suppression and global consistency. The paper was accepted to IROS 2026, and the code and datasets have been released on the project website.
For multiple observations under image-quality constraints in drone-based 3D reconstruction and inspection, where overlap-based planning heuristics cannot guarantee such observations, this work models GSD-constrained k-coverage viewpoint selection as a set multi-cover problem. Candidate viewpoints are generated along outward surface normals, screened by distance, viewing angle, field of view, and occlusion, and then a dynamically weighted greedy heuristic prioritizes coverage gaps and scarce viewing angles. Across two structures, four voxel resolutions, and three coverage requirements, the method achieves full or near-full k-coverage, using only 3.00%–16.39% more viewpoints than ILP.
The authors propose SG^2SLAM, a 3D Gaussian Splatting SLAM method for monocular outdoor navigation that replaces per-frame cross-dimensional correspondence with a chunk-level metric prior, so that tracking and mapping share the same geometric source. The method also introduces a ray-conditioned 3DGS kernel that extends each primitive with a viewing-direction subspace, and jointly optimizes spatial geometry and poses. On the KITTI and Waymo Open datasets, SG^2SLAM outperforms per-frame association baselines in the monocular outdoor navigation task. Code will be made publicly available upon acceptance.
San Francisco-based Bedrock Robotics announced the completion of a $600 million Series C funding round, co-led by Alphabet's growth fund CapitalG and Emergence Capital, with participation from Temasek, Nvidia, Bessemer and others. The post-money valuation is $4.1 billion.
HuaRay Technology exhibited new machine vision and AMR products and solutions at the Electronic Semiconductor Exhibition and Machine Vision Exhibition at Shenzhen Bao'an International Convention and Exhibition Center from October 27 to 29. On the vision side, these included the 65MP CXP large-area-array camera, RayDes camera, 5MP short-wave infrared camera, 16K 256-level TDI line-scan camera, and the R5000Pro and RS3000 code readers, among others; on the AMR side, live demonstrations included the M060C modular chassis, the FCX100B omnidirectional latent fork-lift AMR, and a latent lift AMR for the PCB industry, among others.
Hikrobot has officially launched its "Mingzhi" series of standalone intelligent products, integrating perception, decision-making, and execution into the robot body itself, eliminating reliance on a scheduling architecture so that a single unit can complete the entire material-handling workflow. The series debuts with two products: the Z1-300 underride transport robot and the Z2-1500 forklift transport robot. The company says unboxing, mapping, and task dispatch can be completed in as little as 30 minutes, and a single map can cover up to 200,000 square meters.
A four-wheel-drive autonomous exploration robot based on the Arduino UNO R4 WiFi that can navigate paths and mazes and detect surrounding obstacles. It uses a servo-mounted ultrasonic sensor to perform 180° radar scanning and transmits real-time telemetry over Wi-Fi to an interactive web dashboard, which reconstructs a 2D vector map of the surrounding walls.
The layout drawing says the pallet is centered. The floor disagrees. Our Pallet Mover targets the pallet that actually exists. What is the messiest pickup scenario you have seen a robot handle? #AMR #Robotics
Ropedia released six studies at the first Ropedia Super Week, held October 19 to 23, centered on how first-person human video can be turned into training and execution information usable by robots.
Why it matters: The six studies collectively address how first-person human data can be converted into training and execution information usable by robots, and present cross-embodiment and real-robot comparison results.
Embotech and BMW announced that they have completed the movement of 1 million vehicles using autonomous driving technology at four plants in the UK and Germany, averaging about 3,500 moves per day with availability exceeding 99.5%. Their Automated Driving In-Plant solution places perception, including Hesai lidar, on the production line infrastructure side, so vehicles need no additional sensors, and it is paired with a deterministic safety architecture independently certified by TÜV SÜD.
Xpanner has launched the X1 Piling R6 fully autonomous pile-driving robot, designed to install posts autonomously at utility-scale solar and battery storage sites under the supervision of a single operator. The system is built on the Everstar R6 tracked chassis and uses high-precision GNSS and onboard sensors to navigate autonomously to pile locations. It can drive piles up to 23 feet (7.0 meters) long to a maximum driving depth of 15.7 feet (4.8 meters), and is equipped with a 97 kW engine and a hydraulic hammer with up to 1,260 J of impact energy.
Researchers from Southern University of Science and Technology and other institutions published NI-skins artificial skin in Nature Machine Intelligence; it reads more than 10,000 sensing elements per second and processes signals hierarchically, taking about 23 milliseconds from sensing to interpretation.
Hesai Technology's spatial scanning solution Kosmo has been included by Pixar in a 3D rendering tutorial, where it is used to capture 3D Gaussian splatting data of real-world scenes. Kosmo consists of the handheld capture device Kosmo One and the cloud reconstruction component Kosmo World. Kosmo One integrates a custom lidar, an imaging system of up to 8K, a one-inch CMOS sensor, and a high-precision IMU; the four types of sensors perform microsecond-level time synchronization and factory extrinsic calibration at the hardware layer.
Guangcheng Innovation has launched the KMX series high-precision 4D structured light cameras, aimed at high-precision measurement at the end of embodied intelligence systems. They offer up to 2μm repeat measurement accuracy and up to 60Hz 3D imaging speed, weigh 250g, and are suitable for integration at the end of robotic arms. The series comes with GCI3DLab general vision software and SDK interfaces for real-time 3D measurement and positioning of moving targets, enabling a closed visual feedback loop at the end of robotic arms. Guangcheng Innovation currently offers more than thirty models of 3D structured light products across its Shark, Lynx, and Kite series.
An ECCV'26 paper proposes MV2GF, a multi-view pedestrian detection method that uses the multi-view features and 3D point maps of the pretrained visual geometry foundation model DA3 to aggregate BEV features instead of perspective projection. Under the setting where training and test camera configurations differ, MV2GF achieves MODA on GMVD-D, MVPerception, and Wildtrack that is 4.6, 4.7, and 2.2 points higher than the previous best, respectively.
Beijing Blue Horizon Robotics Technology released its self-developed industrial-grade stereo 3D camera Spatiscan S1 on October 26, with a full-unit price of 13,900 yuan, positioned for measuring transparent, highly light-absorbing, and highly reflective materials and for adaptation to outdoor strong light at the 100,000-Lux level.
FlatVPR proposes a plug-and-play geometric linear residual adapter that applies a residual transformation to features from foundation models such as DINOv2-ViT-S/14 and uses a Pullback Flatness Loss to reduce manifold curvature, so that descriptors between adjacent anchors can be reconstructed by linear interpolation. On the NCLT dataset, the adapter still delivers clear performance gains even when anchor spacing reaches 100 m and severe seasonal changes are present. Map construction is incorporated into an EM framework, split into a continuous M-step for manifold adaptation and an E-step for the anchor selection criterion.
Prateeth Rao and Sachit Rao propose a geometry-aware epipolar correspondence graph method that uses epipolar constraints to guide the construction and pruning of correspondence graphs, then a graph learning module generates a geometric context vector, which is passed through an MLP to output rotation and translation for camera pose estimation.
The authors propose an inspection planning framework that replaces geometric visibility checks with a learned camera information metric based on a classifier variational autoencoder (VAE), to address inspection failures caused by drastic orbital illumination changes. The method implements a guidance, navigation, and control architecture in a photorealistic physics simulator and tests multiple orbital illumination conditions and viewpoints: average brightness reaches 0.57, with 81.6% of observations above 0.25, compared with 0.02 (3.0%) for geometric planning with illumination bands and 0.04 (6.0%) for pure geometric planning.
This study formulates upper-limb functional mobility estimation as an active perception problem, parameterizing reductions in healthy joint limits with impairment coefficients and integrating this into a human model used during robot interaction. The authors propose an online active perception pipeline in which the robot actively probes the user through repeated handover requests while progressively updating a Bayesian belief over the impairment coefficients. Simulation studies and a proof of concept on a robotic platform show that the pipeline can recover the impairment coefficients with relatively high accuracy in real human-robot handovers.
CognitiveReality builds a semantically indexed Gaussian-TSDF map from a robot's RGB-D stream, shared by a VR operator and a tool-calling language agent. A single mapping program adapts to different platforms through configuration alone, accepting poses from robot SLAM, joint kinematics, motion capture, or visual tracking, and handles localization outages with shadow tracking and keyframe-anchored PnP, maintaining open-vocabulary instance identities and per-object quality at 2 Hz.
DGT-Map proposes a self-supervised framework that learns global, directional, and vehicle-conditioned traversability costmaps from RGB-D observations and locomotion signals. It uses a shared multi-task backbone to learn common terrain features across vehicles, while vehicle-specific prediction heads preserve platform-dependent responses. At inference, it outputs a heading-indexed costmap for use by direction-aware planners.
StepTrigger is a contact-state-triggered backdoor attack on VLM-powered legged robots. The trigger signal comes from the foot-ground contact pressure patterns produced when a Unitree Go1 quadruped walks across dense terrain, rather than from prompts or visible markers.
The authors propose VLN on the Fly, an onboard vision-language navigation stack that keeps instruction grounding, planning, and control as separately inspectable stages: a quantized VLM grounds the instruction to a coarse image cell, depth lifts it to a 3D goal, a fast B-spline planner generates a feasible trajectory, and a pretrained reinforcement learning policy tracks the trajectory and outputs motor commands across multiple quadrotors.
The authors propose Haptic Sorter, a unified model-based geometric framework that integrates robotic haptic perception, modeling, and manipulation planning for manipulation when object shape and pose are unknown. The framework uses Bayesian optimization to guide haptic exploration and approximates geometric boundaries with superellipses, describes quasi-static robot-object interaction with an adaptive manipulation potential that encodes object geometry, and proposes an online ordinary differential equation based on model prediction and tactile feedback for real-time pose inference. The system is deployed on a 2D robotic sorting task, varying object geometries to validate the framework's robustness and generalizability in both simulation and a real-world multi-arm setup.
CoMo3R-SLAM is a collaborative monocular dense SLAM system for outdoor multi-agent systems. Centered on learned feed-forward 3D reconstruction priors, it uses dense pointmaps to anchor scale across agents and provides correspondences strong enough to geometrically verify inter-agent links.
The paper proposes the Semantic Autonomy Stack (SAS), a six-layer integration framework that builds semantic reasoning, mission interface, and operational memory layers on a validated ROS 2/Nav2 foundation. It uses a deterministic-first cascade that escalates unresolved or guarded cases to a pretrained vision-language model (VLM).
The authors propose a unified system that simultaneously constrains collision risk along the whole-body trajectory and target visibility during grasping in unknown and dynamic environments, comprising three parts: a velocity-aware gaze policy, a hierarchical subgoal policy, and a receding-horizon whole-body planner. In 400 randomized simulation scenarios and real-world trials on a Fetch mobile manipulator, the system achieved success rates of 70.4% in unknown static environments and 57.9% in dynamic environments, 16.3 and 9.3 percentage points higher than decoupled navigation and manipulation policies, respectively, while significantly reducing collisions.
The special Galilean group SGal(3) is a 10-dimensional Lie group whose important subgroups include the special orthogonal group, the special Euclidean group, and the group of extended poses. This work briefly describes SGal(3) and its Lie algebra, explains how the group structure supports a unified representation of uncertainty in space and time, and points out the group's potential uses for several robotics problems.
The paper proposes an uncertainty evaluation method for ISO 21448 SOTIF analysis: an uncertainty threshold selected under one condition is transferred to a triggering condition, and a test decides at a given error probability whether the performance target is applicable, not applicable, or undetermined under that condition, and provides a lower bound on the number of independent units required for an applicable conclusion.
The authors propose TLP, a point cloud registration method with only one tunable parameter. It uses truncated Lp functions with different p values to assign priority scores to correspondences, select hypotheses, and refine the final model. Targeting the high outlier rates in real-world 3D feature matching, the method uses GPU parallelism to evaluate hypotheses and can on average complete registration from thousands of correspondences within 1 second. On public benchmark datasets, TLP is more accurate than existing single-parameter methods and outperforms two commonly used multi-parameter methods on three of the four evaluated datasets, while maintaining competitive runtime.
Researchers propose an analytical framework based on the OWL perception function for segmenting multiple independently moving rigid bodies during single-camera motion. OWL encodes perceived looming L and perceived rotation ω into a closed-form representation of relative 3D structure, from which the unit direction of relative translation can be analytically recovered per pixel from L and ω alone; points on the same non-rotating rigid body share the same direction, so clustering completes the segmentation. The method is derived in 3D using quaternions, is parallel and camera rotation-invariant, and requires no training data, depth information, or semantic priors.
The authors propose the GEMS framework, which combines a persistent spatiotemporal 3D Gaussian Splatting (3DGS) map with an agentic reasoning pipeline. The map is reconstructed once and thereafter updated only in regions where the scene changes, with those modifications saved. Using this map, the agentic pipeline renders novel views on demand, progressively accumulates evidence to answer spatiotemporal questions, and grounds objects from natural-language queries.
TrySee proposes a modular closed-loop robot task planning pipeline that combines general-purpose VLM reasoning with domain-specific perception and existing robotics components, without requiring VLM fine-tuning or end-to-end demonstrations. The system uses reactive cyclic planning and adds domain-specific modular vision preprocessing to provide online grounding, while a dynamic preprocessing selection mechanism improves scalability.