Skip to content
Only First party#Perception and localization
  1. Beijing Humanoid Robot Innovation Center70

    Beijing Humanoid releases Hi-WBC general whole-body control model on "Huisikaiwu" platform

    Beijing Humanoid released the Hi-WBC general whole-body control model on the "Huisikaiwu" platform on November 2. The model combines human motion references with real physical constraints, enabling robots to adjust how they execute tasks as payload, external forces, and terrain change. It uses a continual skill expansion framework based on MoE and policy distillation, Contact-aware Learning that incorporates contact force prediction, and Perceptive Control that introduces terrain awareness.

    Why it matters: The original text discloses Hi-WBC's technical approach to handling payload, external force, and terrain constraints beyond motion tracking, offering insight into how whole-body control is moving from demonstrations toward real work.

  1. CMU Robotics Institute62

    CMU proposes Wiggle and Go: estimating physical properties of ropes by wiggling to plan manipulation

    A team at CMU's Robotics Institute has proposed the Wiggle and Go system, in which a robot first briefly wiggles a rope or cable while an overhead camera records how the waves propagate; a computer vision algorithm then tracks the motion and estimates the object's physical properties, which are used to plan subsequent actions without repeated trial and error. The model is trained entirely in simulation, and properties are then estimated from the motion of real ropes being wiggled to improve prediction accuracy on real hardware. The research has been accepted to the 2026 Conference on Robot Learning (CoRL) and will be presented next month in Austin; it is partly funded by the U.S. National Science Foundation.

    Why it matters: The method estimates a rope's physical properties from a single wiggle and then plans actions accordingly, offering a transferable approach to deformable object manipulation.

  1. Dexmate72

    Dexmate and CMU Release MeMoREval Benchmark: Robot Storing Camera Frames Beats Text Descriptions

    Dexmate and CMU introduced the MeMoREval benchmark, which uses a two-stage protocol of experiencing first and asking questions later to evaluate what embodied agents should store as memory. It includes 53 tasks and 4 simulated indoor environments, and runs on the Vega 1 Pro dual-arm mobile manipulator. Results show that storing raw camera frames scored 70% and had lower write cost, surpassing text descriptions at 56 to 60%; adding scene graphs achieved the highest overall success rate among all tested models. The protocol also completed two memory retrieval demonstrations on a real Vega 1 Pro using a SLAM map and head camera.

    Why it matters: The benchmark was proposed jointly by the platform developer and CMU; its frame-versus-scene-graph comparison can help with memory design choices, and the real-robot transfer offers a feasibility reference.

  1. RealSense74

    RealSense D585 Pro On-Device Human Detection: Ranging Principles and Safety Architecture Usage

    RealSense explains D585 Pro on-device human detection: the detection model runs on its Gen5 Vision SoC and outputs JSON metadata at 30 fps with class, confidence, pixel box, and millimeter-level distance, without needing a host GPU.

    Why it matters: The article breaks down why on-device detection does not use neural-network ranging and provides a detection distance table for different postures, which can be used directly for safety zone design.

  1. CMU Robotics Institute68

    CMU research team develops human-inspired multisensory fingertips for DeltaHands robotic hand

    CMU researchers designed human-hand-inspired fingertips that can be mounted on the DeltaHands flexible robotic hand. They combine strain gauges to measure pinch pressure with a microphone that captures micro-vibrations through a plastic fingernail, enabling the hand to count paper cups, identify materials, and grasp fragile items such as tofu cubes and potato chips. Each fingertip costs less than 100 US dollars, compared with about 800 US dollars to build a similar setup from off-the-shelf products. The work was presented at the 2026 Intelligent Robots and Systems Conference in Pittsburgh.

    Why it matters: The original text provides details on the multimodal tactile design of the human-inspired fingertips and their cost of less than 100 US dollars, helping readers understand how force and vibration sensing support manipulation control.

  1. RealSense65

    RealSense Perception Studio Min-Z extends stereo depth near-field range to about 2 cm

    RealSense has introduced Min-Z, an improved close-range depth feature in Perception Studio that extends usable stereo depth to about 2 cm on D401/D405 and about 12 cm on D43x/D45x without changing cameras.

    Why it matters: The original explains how Min-Z extends the near-field usable range of stereo depth and the trade-offs between two deployment paths, which can be used to evaluate grasping and close-range perception solutions.

  1. PKU EPIC Lab64

    PKU EPIC Lab proposes UniPart for zero-shot language-driven 3D part segmentation

    PKU EPIC Lab proposes UniPart, a feed-forward cross-modal 3D Transformer conditioned on CLIP text embeddings that uses free-text phrases to select functional parts on point clouds, achieving zero-shot language-driven 3D part segmentation.

    Why it matters: The paper presents a text-driven 3D part segmentation method and a self-built dataset, and validates transfer to real-world language-conditioned grasping.

  1. Boston Dynamics66

    Boston Dynamics releases Spot 5.2 and Orbit update, connecting factory AI data around assets

    Boston Dynamics has released Spot 5.2, positioned as a middle layer connecting the various AI systems in a factory. The Orbit software has been restructured to organize data around assets, can import asset hierarchies and link them to inspections, supports connecting external data sources such as PLC sensors and security cameras, and adds an MCP layer for AI agents to autonomously schedule Spot deployments.

    Why it matters: The original article details the specific changes in Spot 5.2 regarding data connectivity, inspection modalities, and predictive maintenance, allowing readers to assess whether to update their own deployments.

  1. NEURA Robotics63

    NEURA Robotics fully acquires German cleaning robotics company ADLATUS Robotics

    NEURA Robotics announced that its NEURA Mobile Robots GmbH has acquired 100% of the shares of ADLATUS Robotics GmbH, an autonomous cleaning robotics company based in Ulm, Germany.

    Why it matters: The original article provides details on the acquired company's existing equipment, navigation software, and plans to integrate with Neuraverse, which helps explain NEURA's cleaning business strategy.

  1. 北京智在无界科技有限公司64

    BeingBeyond releases Being-H0.8, the first implicit world action model to introduce touch in pretraining

    BeingBeyond has released Being-H0.8, which it calls the world's first implicit tactile world action model, unifying vision, touch, action, and future state changes into a single latent space, with pretraining on more than 500,000 hours of first-person human video.

    Why it matters: The model brings touch into large-scale pretraining along with a self-developed pseudo-labeling system, letting readers understand the path by which tactile data goes from a lab signal to a scalable pretraining resource.

  1. Lavender Hill Capital Partners63

    Tashan Technology Completes Series B Round Worth Hundreds of Millions of Yuan, Building an End-to-End Tactile Sensing Industry Layout

    Tashan Technology recently announced the completion of a Series B financing round worth hundreds of millions of yuan, with investors including Lavender Hill Capital Partners (LHCP).

    Why it matters: Operating data of tens of thousands of units shipped per month and a share exceeding 80% lets readers gauge the actual progress of tactile sensing toward scaled deployment.

  1. 1X Technologies65

    1X releases 25-DOF tendon-driven dexterous hand for NEO humanoid robot

    1X Technologies announced a 25-DOF tendon-driven dexterous hand for the NEO humanoid robot platform, which will ship with every NEO unit.

    Why it matters: The original article provides key specifications and production progress for the 25-DOF tendon-driven hand, which readers can use to assess its actual impact on humanoid manipulation capabilities.

  1. PKU EPIC Lab68

    WAM-TTT: Steering World-Action Models at Test Time by Watching Human Videos

    PKU EPIC Lab proposes WAM-TTT, a test-time training framework that steers world-action models from raw human videos without robot actions, human-side annotations, or task-specific fine-tuning. The method absorbs human videos into a lightweight adaptive memory inside a frozen WAM through self-supervised video prediction, and uses a meta-training stage with paired human-robot data and a key-value memory reconstruction objective to align human demonstrations with robot behaviors. Experiments show it consistently outperforms in-context human-video conditioning baselines across diverse manipulation tasks and generalization settings.

    Why it matters: The original work proposes a method for adapting a frozen world-action model at test time using only unlabeled human videos, without robot demonstrations or fine-tuning, offering methodological reference value for researchers who want to reuse it.

  1. 蚂蚁灵波科技 · YouTube61

    LingBot-Depth 2.0 spatial perception model for robots released, can see through glass and mirrors

    LingBot-Depth 2.0 has been released, a next-generation spatial perception model for robots that focuses on perception of glass, mirrors, and transparent objects. It was trained on 150M-scale data (50 times that of v1.0) and took first place in 12 of 16 depth completion benchmarks; on the hardest indoor scenes, depth error RMSE dropped from 0.132 to 0.062.

    Why it matters: The original text provides benchmark rankings, the change in depth error on the hardest indoor scenes, and an open-weights link, helping readers judge the improvement over the previous generation and how reusable it is.

  1. Dexterity66

    Dexterity speeds up Foresight visual perception 17x with NVIDIA, from 1508ms to 90ms

    Dexterity says that after NVIDIA acceleration, the visual perception pipeline of its world model Foresight cut per-cycle latency from 1508ms to 90ms, a 17x speedup.

    Why it matters: The original breaks down latency data and optimization methods for each stage of the pipeline, giving robotics developers a GPU acceleration approach they can adapt to their own perception systems.

  1. Figure AI82

    Figure AI Releases Helix 02: A Full-Body Autonomous Control Model for Humanoid Robots

    Figure AI has released Helix 02, extending control from the upper body to the entire humanoid robot and enabling continuous, integrated autonomous tasks combining locomotion and manipulation. It adds a new System 0 whole-body controller, trained on over 1,000 hours of human motion data and simulation-based reinforcement learning, which outputs joint commands at 1 kHz via a 10M-parameter network, replacing roughly 109,504 lines of hand-written C++ code; System 1 connects head-mounted vision, palm cameras, fingertip touch sensing, and proprioception, outputting whole-body joint targets at 200 Hz.

    Why it matters: The original text provides details on Helix 02's architectural layers, training data, and demonstration tasks, allowing readers to understand the specific implementation approach for full-body autonomous control.

  1. Sharpa63

    Sharpa announces mass production of SharpaWave dexterous hand, to exhibit at CES 2026

    Sharpa announced that mass production of its flagship dexterous hand, SharpaWave, has begun, with early customers including global technology companies and research universities.

    Why it matters: The manufacturer announced that the dexterous hand has entered mass production and provided technical details such as degrees of freedom, tactile precision, and simulation platform compatibility, which readers can use to assess its research usability.

  1. Figure AI67

    Figure AI releases third-generation humanoid robot Figure 03

    Figure AI has released its third-generation humanoid robot, Figure 03, redesigned from the ground up for the Helix VLA model, the home, and large-scale mass production, covering both software and hardware.

    Why it matters: The official announcement fully explains the new-generation humanoid robot's perception, touch sensing, home-safety design, and mass-production plans, so readers can understand its hardware changes and production capacity targets.

  1. LimX Dynamics66

    GVF-TAPE Algorithm from LimX Dynamics and Partners Accepted to CoRL 2025, Enabling Robots to Learn Manipulation from Video

    The paper "Generative Visual Foresight Meets Task-Agnostic Pose Estimation in Robotic Table-top Manipulation," a collaboration between SUSTech CLEAR Lab, LimX Dynamics, and the University of Hong Kong, has been accepted to CoRL 2025.

    Why it matters: This research presents a concrete technical approach and experimental figures for learning manipulation from human video, helping readers understand the feasibility of training robot manipulation with video data.

  1. 线性资本72

    Bot Auto completes first fully driverless truck real-world road test

    Linear Capital portfolio company Bot Auto announced it has completed its first real-world road test in which the truck truly performed all operations on its own, with no one in the driver's seat, the back seat, or at a remote-control terminal, and no escort vehicle. During the test, the truck on a Texas highway handled scenarios including avoidance, detouring, merging, and responding to the appearance of a fire truck. The vehicle is equipped with 15 cameras, 8 lidars, and 3 millimeter-wave radars, with redundancy in perception, computing, actuation, and power supply; after the primary server is shut down it can switch to a backup server, and if the backup is also shut down, the vehicle control unit takes over to pull over and stop.

    Why it matters: The article sorts out the differences between Bot Auto's truly driverless road test and earlier modes with safety drivers, escort vehicles, and remote control, helping readers understand the differences in its technical approach.