MiTaS is a multi-resolution tactile representation framework that uses modality-specific convolutional stems and a Transformer-based fusion structure to integrate information from an RGB camera, a GelSight Mini vision-based tactile sensor, and a high-frequency event-based Evetac sensor, then conditions a flow-matching policy to perform downstream tasks.
Move-Then-Operate is a vision-language-action (VLA) framework that explicitly decouples robotic manipulation into two behavioral phases: coarse relocation (move) and contact-critical interaction (operate), using a dual-expert policy routed by a learnable phase selector.
WARPED is a framework that synthesizes realistic wrist-view observations and actions from human demonstration videos, enabling visuomotor policy training with only monocular RGB data. The system uses vision foundation models to initialize the interactive scene, employs a hand-object interaction pipeline to track the hand and manipulated object and retarget trajectories to a robotic end-effector, and finally uses Gaussian Splatting to synthesize photorealistic wrist-view observations for directly training policies.
Safe-Night VLA is a multimodal manipulation framework that integrates long-wave infrared thermal perception into a pre-trained vision-language backbone, enabling robots to perform semantic reasoning based on thermodynamic properties and to enforce deterministic workspace constraints during policy execution through a control barrier functions safety filter.
AFFORD2ACT (Affordance-Guided Keypoints to Act) studies how much manipulation capability can be retained when only sparse 2D keypoints are used as input: an off-the-shelf language-conditioned affordance model locates task-relevant regions, selects up to 15 object keypoints plus 4 end-effector keypoints, compresses the policy input to 19 2D keypoints (38 coordinates), and then uses lightweight gating to decide which keypoints to focus on during execution.
A study proposes introducing a laser pointer into a vision-based portable pipeline for target disambiguation in multi-object scenes, to support grasp selection and preshaping in upper-limb prostheses. The method uses a mixed reality strategy to reuse existing high-quality segmentation masks to generate training images and retrains a deep convolutional neural network. Experimental results show target disambiguation accuracy improves by up to 9% over existing solutions while computational cost remains unchanged.
CW-WAM is a world action model that jointly predicts video, motion, and contact wrench. Inspired by hybrid motion-force control, it uses a unified interaction module to predict time-aligned motion-wrench chunks as control references. It structures motion-wrench dependencies with pair-causal attention and retains past interaction information in a fixed-length wrench memory; the predicted wrench chunk serves as the force reference for a 20 Hz outer-loop admittance controller.
The paper proposes AtomicRL, a framework that uses atomic task structure to improve reinforcement learning post-training of vision-language-action (VLA) policies. Its core components include Semantic-Aware Policy Optimization (SAPO), which uses semantic gating to prevent TD residuals from failed atomic subtasks from propagating across tasks, thereby mitigating credit contamination; and Bottleneck-Adaptive Target Exploration (BATE), which dynamically coordinates task sampling and local exploration incentives based on the mastery level of each subtask.
The study proposes a unified closed-loop manipulation evaluation framework that assesses frontier AI models' ability to control a robot directly through visual observations, robot proprioception, and structured tool calls on the LIBERO-10 benchmark, without task-specific fine-tuning.
DONIC is the first generalist bimanual dexterous tracker jointly trained across manipulation categories and human motion sources. It is released alongside BiHOI, a multi-view dataset containing 1,575 sequences covering 229 objects, in which reconstructed interactions are retargeted into robot references.
SpatialDex is a framework that learns 3D-grounded bimanual dexterous manipulation from RGB-only human videos through reconstruction-augmented visuomotor pre-training, learning human behavior together with its spatial context. It reconstructs scene geometry and estimates hand motion from each video, aligns the two at a video-level scale using the hand-size prior of 3D hand estimates, and uses the reconstructed geometry as both a policy input and a prediction target.
FlowVLA separates action generation from physical-completion estimation. A shared π0.5 action expert learns approach, pour, and return, while a target-independent estimator fuses visual, joint, wrist-wrench, and recurrent force-belief histories to estimate transferred mass. It triggers return when this estimate plus predicted residual outflow reaches the requested mass.
The paper proposes SERA (Self-Evolving Robotic Agent), which organizes a high-level VLM as a robotic agent with an editable harness, paired with a fixed VLA executor for low-level control, to handle non-Markovian dependencies in long-horizon manipulation.
The paper proposes PAL (Plan first, Act Later), a hierarchical memory framework for vision-language robot manipulation. A high-level vision-language model (VLM) converts streaming visual observations into a subtask_list, and at execution time selects the currently relevant subtask from it to pass to a low-level vision-language-action (VLA) policy.
The paper proposes Relational Geometry Forcing (RGF), which injects geometric priors from a frozen 3D foundation model into World Action Models (WAMs) by aligning pairwise feature relations instead of pointwise feature alignment, avoiding the degradation of action learning caused by dense geometric targets over-constraining compact action features.
The authors propose TIDE (Training In DEployment), a human-in-the-loop online post-training system that uses a provenance-conditioned supervised learning framework to update the policy online from human interventions and policy rollouts during deployment, without learning a critic, reward model, or action scorer.
The authors propose DepState-JEPA, a dual-stream RGB-D joint-embedding predictive architecture that connects spatial understanding with action-conditioned prediction through a shared latent state, DepState. Built on V-JEPA 2.1, the framework uses depth both as a complementary observation stream and as the grounding for metric 3D space, and constrains state learning with ray-conditioned readouts, object-trajectory supervision, and calibrated forward kinematics.
ACT-Flow is an online learning framework that combines an adaptive Action Chunking Transformer (ACT) with a conditional residual flow policy, allowing ACT to absorb successfully executed actions through behavioral cloning, with residual targets recomputed relative to the updated prior.
PHD-Hand is a linkage-driven humanoid dexterous hand with three drive degrees of actuation (DOA) per finger. It uses a unique flexure hinge and elastic drive rods made of spring steel to simplify the transmission, avoiding the need for many miniature moving parts, and can be manufactured with budget-level 3D printers, lowering the barrier to use and development of dexterous hands and bringing its size closer to that of a human hand. A prototype has been used for grasping and object manipulation evaluation, and the paper also presents the latest research results on the hand's kinematics and control. It has been accepted by IEEE-RAS Humanoids 2026.
COWBOY is a scalable sim-to-real framework for learning contextual whole-body manipulation. On whole-body dexterous grasping tasks, it transfers zero-shot to hundreds of objects and diverse in-the-wild scenes, including cluttered tabletops, shelves, drawers, and other unstructured environments.
T-WLA is a World-Language-Action architecture that uses shared visual memory to guide both action generation and tactile correction, addressing the problem in contact-rich manipulation where contact conditions change before the visual policy updates.
The paper presents DONIC, a generalist bimanual dexterous tracker jointly trained across manipulation categories and human motion sources, and releases BiHOI, a multi-view dataset containing 1,575 sequences covering 229 objects, emphasizing sustained bimanual interaction and retargeting reconstructed interactions into robot references.
UniPicker Pro: strength and dexterity combined. From tiny, delicate objects to demanding loads, UniPicker Pro brings speed, precision, and power to robotic grasping. Learn more: https://www.unix-group.ai. Contact us: global@unix-group.ai. #UniXAI #UniPickerPro #RoboticGripper
Echo proposes a framework that treats two task status signals, stop and failure, as two dimensions of each action chunk and denoises them together with robot actions, thereby distinguishing ongoing execution, successful completion, and failure. The method requires no frame-level progress annotations; it provides complementary supervision by constructing successful, failed, and recovery trajectories, and it lets status signals denoise faster than actions so that subsequent action denoising steps can build on cleaner status estimates. Real-world manipulation experiments show that Echo accurately predicts task status and consistently improves policy performance.
SEER Robotics demonstrated its robotic arm completing the full coffee-making process without human intervention, from picking up the portafilter, receiving grounds, distributing and tamping the grounds, locking the portafilter, extraction, to cleaning. This is the longest multi-step robotic coffee task the company has demonstrated to date.
The Everbot third-generation robot completed zero-shot grasping of a button and placing it into a tray on real hardware, taking 58 minutes in total (video sped up 175 times); inference, recognition, and movement all ran onboard a Jetson Orin Nano Super, without relying on the cloud.
Brightpick CEO Jan Zizka explained how warehouse robots request human assistance when picking fails, and said the system has multiple layers of fallback mechanisms for unexpected situations. He also described receiving an injunction obtained by Ocado about 12–13 hours before the LogiMAT trade show, which prevented Brightpick from demonstrating the Gridpicker at the show; Zizka discussed the disruption the injunction caused to the launch event.
The author has open-sourced the hand exoskeleton TAKTO ONE, designed during their master's thesis, with CAD, PCB, and firmware all free to download, totaling 29 parts. It can directly read 12 finger joints and is used to collect hand data to train robot hands or drive robotic hands. The author also released a TAKTO Lite version without motors; because motors account for about 85% of the parts cost, the Lite version costs about 164 euros plus circuit board costs, and the author is openly soliciting designs for a thumb module.
A Reddit user designed a cookie pick-and-place line made up of 12 KUKA delta robots that place cookies into trays, and is currently running into problems with failed picks, crumbs and grease, and cycle times that are too slow. Some cookies cannot be picked up because of insufficient airflow, and the user is considering increasing the cylinder diameter of the gripping system; when the suction cups pick up cookies, they bring up crumbs and grease, but switching to a hand-like gripper would crush adjacent cookies on the conveyor. The line's programmer says the tray conveyor and robot speeds need to be increased in sync to shorten the waiting time for placing, but the user believes this approach cannot solve the existing problems.
Agile Robots, Audi, Dassault Systèmes and NVIDIA announced a proof-of-concept collaboration at NVIDIA GTC Berlin to use mobile humanoid robots for parts picking and sequencing at Audi's Ingolstadt plant for final vehicle assembly.
Why it matters: The original text explains the validation process in which a virtual twin precedes physical deployment, letting readers understand the practical path for humanoid robots entering automotive logistics.
A robotics project uses evolutionary search and AI-generated CAD in simulation to let a robot iteratively design and test tools such as grippers to improve task performance; the robot's motions still come from a predefined task controller. The next step is to validate on real hardware using an Arctos robot and a 3D printer, and the author is seeking suggestions for automatically removing gripper fingers from the print bed.
Three days after Geek+ invested 50 million yuan in Shenyi Robot for a 20% stake, Shenyi Robot, which was founded just nine months ago, unveiled its self-developed embodied foundation model SE-1 at the 2026 CIIF, along with Sculptor, a multi-task embodied operation robot focused on heavy-load logistics, and Edge, a dual-arm robot focused on fine industrial operations.
US embodied intelligence startup Eka Robotics announced the close of a $150 million Series A round at a post-money valuation of about $1.45 billion, co-led by a16z and F-Prime, with Samsung Next and others participating.
The number of patents related to flexible continuum robotic arms in China is continuing to increase, including "An Origami Bionic Dual-Segment Flexible Continuum Robotic Arm" and "Flexible Continuum Robotic Arm with Adjustable Stiffness, Stiffness Adjustment Device, and Stiffness Adjustment Method," with major participants including Wanxun Technology, Qingyu Technology, and Shiyuan Co., Ltd.
CondenseVLA proposes adding a learnable history compression module to the decoding side of VLA policies. It uses learnable queries and multi-head cross-attention to distill the per-frame features cached by the VLM backbone into a fixed number of slot tokens, and applies auxiliary regularization that penalizes overlap among slot attention maps and decorrelates slot embeddings to reduce redundancy.
The EMMA framework enhances vision-language-action (VLA) policies through a generative data engine and training pipeline, addressing the high cost of collecting robot manipulation data across varied appearances. Its core component, DreamTransfer, is based on a diffusion Transformer and generates multi-view-consistent, geometrically plausible manipulation videos, supporting text-driven editing of foreground, background, and lighting while preserving 3D structure.
ForeTac is a predictive contact guidance framework for contact-rich manipulation that uses a represent-predict-evaluate-guide loop: it encodes the current tactile state, predicts future tactile latent trajectories conditioned on candidate action sequences, and evaluates each trajectory with a contact-quality objective learned from successful and failed physical interactions. At inference, gradients of this objective are backpropagated through the tactile predictor to guide action generation before execution, while all model parameters remain frozen.
The paper proposes an end-to-end pipeline that uses an instrument-state logger mounted on the shaft of a standard laparoscopic instrument (combining inertial, time-of-flight, and Hall sensors, with no external camera or tracker) to capture the instrument pose and jaw state during a surgeon's operation, which is then used to train a surgical robot policy and validated on live animals.
The paper proposes a modular method that autonomously collects data in local segments of contact-rich tasks and uses offline reinforcement learning to train local policies, which are then integrated with traditional motion planning to complete long-horizon assembly tasks. Across four real-world tasks, using only 2-2.5 hours of autonomous data collection, the method achieved an average success rate of 96% on long-horizon tasks, compared with 55% for a planning-only baseline and 37% for the strongest end-to-end baseline; the local policies also reduced applied force by 49%. Performance remained at 90% in scene-level out-of-distribution scenarios, with more than 1,500 real-world evaluation rollouts completed in total.
PAKE proposes a whole-body loco-manipulation control framework that decomposes the problem into partial reference motion generation and low-level imitation control. It introduces a Kinematic Normalizing Flow (KNF) model trained on a large-scale motion generation dataset to generate diverse partial motion references; a high-level controller navigates the KNF latent space to exploit redundant solutions, while a low-level controller learns dynamically consistent motion execution.