Robotic Manipulation
Overview
Robotic manipulation studies how machines use sensing and motion to change objects and environments, from picking and assembly to tool use and handling deformable materials. Approaches combine geometric planning and force or impedance control with imitation learning, reinforcement learning, diffusion policies, and vision-language-action models; world models, memory, tactile feedback, and human-video demonstrations are also used. Methods are tested in simulation and on real robots using suites such as LIBERO, RoboTwin, ManiSkill, and RLBench, across clutter, bimanual coordination, mobile manipulation, and precision contact tasks. A central challenge is turning benchmark performance into safe, reliable long-horizon behavior across new objects, cameras, robot bodies, and real-world dynamics, while limiting data collection and inference costs.
Latest papers
1,805 papers244 papers in the last four weeks, down 37% on the four weeks before. 1.7% of all new papers.
When Does Cross-View Consistency Matter in Multi-View Appearance Augmentation? A Controlled Study for Visuomotor Policies
ACE-CaP-0: Real-World Code-as-Policies in the Era of Experience
STVLA: Spatial-aware and Temporal-memory forVision-Language-Action Models
JustRUA: A Controlled Study of GPT-6 Robot-Use Agents
GraSLA: A Lightweight Spatial-Language-Action Model with Grasp Hypotheses as Geometric Priors
Generalized Inverse Kinematics via Conditional Flow
PolyMorphWBC: Whole-Body Control from End-Effector Trajectories for Quadrupedal Mobile Manipulation across Arm Morphologies with Extended Reach
CrowdMAPS: Crowdsourcing Multiple Actions per State for Real-World Imitation Learning
Fast Inference-Time Policy Refinement via Action-Affine World Models
Learning Low-Latency Action Verification for Frozen Vision-Language-Action Models
From Prediction to Action: Learning Physical Context for Deformable Linear Object Manipulation
DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
Force-Closure Condition for Cooperative Multi-robot Pushing of Rectangular Objects
Learning Heterogeneous Tactile Representation for General Dexterous Manipulation
PCTac: Mechanochromic Photonic-Crystal Tactile Sensor for Dexterous Manipulation
GAP: Generalized Adaptation Prior for Dexterous Piano Playing
SPARK: Simple Post-training for Adapting pRetrained Knowledge to Robot Control
Thermo-VLA: A Thermal-Aware Vision-Language-Action Model with Gated Residual Modulation for Robotic Manipulation
Decoupled Asynchronous System for Humanoids (DASH): Latency-Agnostic Scaling of Vision-Language-Action Models for Humanoid Manipulation
PAIR: Proprioception-Aligned Image-Ray Representations for Viewpoint-Robust Manipulation
HolidayFactory 1.0: Simulating, Generating, and Evaluating Industrial Robot Tasks
Qubit-to-Motion: Traceable Action Provenance in Hybrid QRL for Robotic Bin-Packing Sequencing
Qubit-to-Motion, an evidence-centered framework for support-constrained sequencing and simulated robotic placement. Four Pauli- readouts provide the sole learned input to the actor and critic, while a support mask enforces admissible object selection. Each decision is recorded and linked via step indices and SHA-256 fingerprints to a sealed route, deterministic Cartesian and inverse-kinematics waypoints, prescribed joint motion, and MuJoCo contact replay. A frozen checkpoint is evaluated across five deterministic 30-object scene variants. All 150 placements satisfy support, containment, and settling criteria. A fixed-state zero-readout intervention changes of decisions, demonstrating dependence on the joint circuit representation. Mean macro travel is ~m, inference requires