The study proposes CPG-ENN, a quadruped locomotion framework that uses a symmetry-based eight-neuron central pattern generator (CPG) to supply gait rhythms, while a pretrained execution neural network (ENN) shared across legs maps rhythmic signals and trajectory parameters to joint-angle references without online inverse kinematics.
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.
This study proposes using a stochastic motion tokenizer to generate concise discrete motion representations that can be fed directly into a whole-body tracking policy to execute dance, fighting, and walking motions on the Unitree G1. At matched capacity, the tokenizer reconstructs Unitree G1 motions better than FSQ and VQ-VAE. In real-robot demonstrations, each generated motion lasted 16.8 seconds, and a 35-second fighting-to-dance sequence was completed.
The study proposes extending PPO into phase-conditioned policy optimization, using a phase-conditioned critic, phase-gated rewards, and a multi-phase reset curriculum to train a single policy for long-horizon loco-manipulation tasks. The method adopts a decoupled hierarchical control architecture that separates low-level balancing dynamics from high-level phase execution, and validates multi-stage skills on real hardware on the Unitree G1 and a Boston Dynamics Spot equipped with a manipulator arm, while testing design choices through ablation experiments.
Wouf, the quadruped robot in the Cubic Doggo project, managed to stagger a few steps while carrying a 1 kg load (about half its own weight), even though the steering gait code had not been fixed. Its ST-3215-C018 servos are not fast, and the author is excited about the result of this payload test and is considering having it carry the SO-ARM101, which also uses 12 servos. The build code has been released on GitHub.
AgiBot has released the safety-certified heavy-duty embodied industrial robot Genie G2 Max, featuring a 20 kg single-arm payload and 40 kg dual-arm payload, with dual-arm peak payload up to 50 kg and repeat positioning accuracy of ≤±0.5 mm, targeting depalletizing and palletizing, material transfer, and production line feeding.
RoboParty has released the RP1 full-stack open-source humanoid robot, opening up complete-robot and joint hardware documentation, the PartyOS software foundation, and a pipeline from simulation training to real-robot deployment. The RP1 features ROMOMO's self-developed joint modules; the 100-diameter module has a measured peak torque of up to 160+ N·m, and with whole-body control algorithms it can walk, run, jump, negotiate obstacles, and get up after a fall. The platform also offers the UFO motion-control framework based on unsupervised reinforcement learning, dual-battery hot swapping, and a modular, maintainable structure, and it has launched a closed beta co-creation program.
Anno Robotics' "Anno AI Robot Latte Art Coffee Printing Bar" won the Gold Prize at the 2026 Canton Fair Design Innovation Award (CF Award). A total of 2,112 companies and 4,218 products were entered, and only 222 won awards. The product uses a self-developed six-axis robotic arm and motion control system to complete the full process from grinding, extraction, and milk frothing to latte art and printing, producing a cup in 90 seconds with repeat positioning accuracy of ±0.03mm, and it connects to Alibaba's Qwen large model to support one-click AIGC image generation. The machine occupies less than 2㎡, supports 24-hour unattended operation, and has entered more than 100 cities in China and been sold to more than 80 countries and regions.
The paper proposes a path-tracking algorithm for articulated vehicles with a single trailer. It applies a lateral control law at the trailer level and makes short-horizon predictive adjustments to the tractor steering angle, so that both forward and reverse motion converge stably to the target path. The method is geometry-based and requires no per-vehicle calibration or training. In a high-fidelity physics simulation, the steady-state or final lateral error on straight-line, circular-arc, and reverse docking trajectories was 0.02–0.05 m. The authors tested it on a small-scale tractor-trailer robot, with the controller running on a low-cost microcontroller; under about 200 ms of pose-detection latency, the lateral error in repeated reverse docking was below 13 mm.
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.
X-WBC is a cross-embodiment humanoid whole-body control foundation framework that separates relatively shared human motion semantics from embodiment-specific physical execution. It uses human-centered command tokens to align full human motion, robot reference motion, and sparse VR observations; a causal Transformer learns reusable temporal structure from mixed multi-robot rollouts, and lightweight robot-specific modules map the shared representation to each robot's proprioception and action space.
Why it matters: The paper proposes a cross-embodiment humanoid whole-body control framework that decouples shared human motion semantics from each embodiment's execution, and validates joint training on nine simulated embodiments and four real robots.
This work proposes using sample-based model predictive control (SMPC) in simulation as an automated expert to generate large-scale offline datasets, then training an off-policy reinforcement learning agent with purely sparse task rewards, eliminating the need for manual dense reward design. Combined with a low-level dynamic stability controller, the high-level agent's learned policies can exceed the original optimal control teacher on task objectives. The authors completed sim-to-real deployment on an arm-equipped Spot quadruped and a G1 humanoid; the paper is the accepted version for CoRL 2026.
The study proposes DSMS, a contact-implicit, direct-simulation-based multiple shooting framework that converts kinematically feasible reference motions into dynamically feasible whole-body trajectories. It handles contact, friction, collision, and joint limits within a nonlinear program without presetting a contact schedule. Compared with existing retargeting methods, DSMS accelerates motion-imitation reinforcement learning training and yields policies with high success rates and low tracking error. The authors achieved zero-shot sim-to-real transfer on the Unitree G1, completing walking, jogging, contact-rich crawling, and a 180-degree jump-turn.
The study proposes Staged Multi-Agent Training (SMAT), a method that jointly trains a musculoskeletal human agent and a bilateral hip exoskeleton control policy in physics simulation to learn an assistance policy that accounts for human-device co-adaptation. The policy was transferred to a physical hip exoskeleton without subject-specific retraining and tested in eight healthy adults; indirect calorimetry showed that active assistance reduced net metabolic rate by 19.7% relative to the passive device (p < 0.001), the hip mechanical positive-power ratio was 0.984, and effective assistance was maintained at walking speeds and terrains beyond the training conditions.
The paper proposes a robot-agnostic compliant control framework that separates ROS controller boilerplate from control-law implementation through a plugin-based execution architecture, extending the ROS control ecosystem. The framework uses Pinocchio to compute robot kinematics and dynamics from URDF models, and its standardized joint and Cartesian command interfaces support motion references, stiffness and damping gains, nullspace targets, desired interaction wrenches, and independent feedforward inputs.
This paper proposes a ZMP modeling method for quadruped robots carrying payloads via a passive arm, relating the stiffness, damping, and payload mass of the passive interface to the ZMP response, and notes that lightly damped passive modes may overlap with gait harmonics.
This study proposes a spatiotemporal informative planning method based on model predictive path integral (MPPI) control for unmanned surface vehicles (ASVs) to search for and capture multiple drifting targets. The planner couples target drift prediction with exploration objectives through a cost design, optimizes continuous trajectories over long horizons, and directly generates kinematic commands; in the interception stage, pure pursuit guidance is used to achieve physical capture. In simulation, the method captured 18% more targets than the strongest baseline, and the approach was validated in real-boat trials.
The authors propose Crazyflow, a differentiable drone simulator built on JAX that combines high-fidelity dynamics with GPU acceleration, scaling to millions of environments or swarms of thousands of drones on a single CPU and GPU. Real-world experiments show that it supports analytical-gradient policy learning, achieving high-accuracy tracking without domain randomization, and sampling-based obstacle avoidance at over 500 million steps per second. The authors also demonstrate in-flight reinforcement learning: after tossing a physical drone into the air, training a recovery policy from scratch took only 0.37 seconds, and the drone was caught before landing.
PrepTurn is a reinforcement learning framework for anticipatory maneuvering in continuous local navigation. It maps the current and next pose goals directly to joint actions, letting the upcoming goal influence how the robot approaches the current one. The framework uses a persistent-credit reward to encourage completing the entire goal sequence while in motion, and an automatic curriculum that gradually increases goal difficulty to stabilize turning behavior; simulation and onboard Unitree Go2 experiments show high sequential traversal completion rates and reduced stop-and-go motion.
The authors propose dynamic window pure pursuit (DWPP), which moves command velocity computation into the v-ω velocity space and selects, within the dynamic window, the point closest to the line ω=κv, thereby explicitly incorporating velocity and acceleration constraints. Experiments show that DWPP avoids constraint-violating commands and achieves better path-tracking accuracy than conventional pure pursuit methods. The method has been integrated into the official Nav2 repository and made publicly available, and the paper has been accepted by Robotics and Autonomous Systems.
This survey constructs a closed-loop task-execution architecture for collaborative continuum robots (CCRs) from an automation-system perspective, explaining how target-task specifications are transformed into coordinated execution by multiple continuum robots. Based on this, the authors classify CCRs into three modes: separated collaboration, assistance collaboration, and parallel collaboration. They then review research progress in terms of multi-robot task allocation, structural design, modeling, motion planning, control strategies, and sensing mechanisms, and finally discuss key technical opportunities and the system-integration challenges of real-world deployment.
The paper proposes using virtual nonholonomic constraints (VNHCs) to address the problem of nonprehensile robotic catching, designing and controlling the motion of the end-effector directly based on the object's state before it is caught, without the need to predict the object's trajectory and generate a time-varying robot trajectory. In the planar case, the method can catch the object at a desired height and minimize the relative normal velocity at the instant of contact; the same VNHC function depends only on the target catching height and can cover a range of initial conditions. After contact, model predictive control (MPC) is used to bring the object and end-effector to rest at a specified position. The approach has so far been validated in simulation.
This project explores controller design and trajectory generation for a quadrotor. It uses a differential flatness approach to generate minimum-snap trajectories based on waypoint navigation, and a simple PD controller to track the trajectories. Simulation results show that the method can generate smooth trajectories for the quadrotor's position and velocity.
For a quadrotor equipped with a two-DOF robotic arm that can physically interact with its environment, this study implemented and compared two controllers: Cartesian impedance control and adaptive sliding mode control. Under bounded disturbances, both controllers performed a simple hovering task with the same initial conditions and target position, and were validated based on a full dynamics model.
The research team proposes the FLORE framework, which uses two force-sensing UMI-style handheld grippers and wearable trackers to collect robot-free demonstrations, and trains a force-reactive diffusion policy that takes vision, proprioception, and measured force wrenches as input and outputs actions and per-hand force commands. A companion hybrid compliant whole-body controller uses integral force error from real force/torque sensors to achieve single-hand force tracking while maintaining balance. In real-robot tests on three contact-rich tasks, the force-reactive policy achieved an 80% success rate, compared with 26.7% for a position-only policy under the same controller.
This paper proposes a smooth analytic gradient for frictional contact that replaces the exact contact sensitivity only in the backward pass, leaving the forward hard-contact solve and the hardware-feasible trajectory unchanged. A policy trained on 100,000 samples in under 2 minutes on a desktop PC transfers to hardware without fine-tuning; under a trotting policy the robot withstands pushes and traverses gravel and slippery ground. In trajectory optimization, the gradient finds contact-rich solutions that the exact gradient and the Signorini relaxation gradient miss.
The research proposes a real-time motion planning framework that retains the full SE(2) model of a quadruped robot along with its anisotropic, heading-coupled velocity constraints, and accepts the resulting nonconvexity. The method parameterizes trajectories as composite Bernstein polynomials, uses the convex hull property to prove that obstacle avoidance and input constraints hold throughout the entire planning horizon, and runs in closed loop at online replanning rates, sending body velocities to the motion controller.
The study proposes a distributed predictive flocking controller in which each miniature vibration-driven robot independently solves a nonlinear model predictive control (NMPC) problem and exchanges only predicted trajectories with local neighbors, comparing six barrier and control barrier function (CBF) safety formulations on a digital twin of an open-source vibration-driven bristlebot in a 0.64-square-meter arena.
The paper proposes RoKON, a data-driven predictive control method for robotic manipulation that does not require learning an explicit dynamics model. The method uses a Gaussian-kernel nonlinear predictor to represent the input-output data relationship, proposes a damped Gauss-Newton method that exploits the least-squares structure of the control problem to achieve real-time control, and introduces a hybrid buffer management strategy that incorporates informative online data while retaining relevant offline data, allowing adaptation to changes in dynamics.
The paper proposes the VisGeo-PipeTrack servo control framework, which uses YOLOv5 to detect pipe labels and combines visual geometric constraints for localization, enabling UAVs to track pipes with dynamic motion such as vibration. On the control side, it adopts a model predictive control-based real-time corrected adaptive PID (MPC-RCAPID), which estimates the model online via recursive least squares and adjusts PID parameters through rolling optimization.
The study compares cascade PID, hover LQR, BL-SMC, linear MPC with LQR fallback (LMPC-F), a preview-enabled LMPC-F, and a geometric SO(3) controller on a nonlinear plant with rotor gyroscopic coupling, motor lag, and saturation. It finds that a unified twelve-candidate tuning protocol changed the mean-RMSE ranking obtained under fixed baseline settings, moving BL-SMC from last to first; the new ranking recurred on 50 fresh realizations and remained unchanged under a 48-candidate budget.
Zero-HLM proposes a real-to-sim-to-real framework that turns smartphone captures and simulated source demonstrations into scalable training data for humanoid loco-manipulation. The study generated 4,800 physically validated demonstrations and achieved zero-shot transfer on three Unitree G1 tasks; GR00T N1.7 trained in simulation reached a mean full-task success rate of 66.7%, versus 46.7% with real-data training.
The study proposes a two-stage Region of Attraction (RoA) certification procedure: contraction-based stability analysis first determines a target set, which Taylor-model reachability analysis then verifies, thereby constructing a context-dependent safe RoA at runtime and enabling safe switching among high-assurance controller candidates.
LECO converts online trajectory optimization for dynamic target tracking into offline learning of a direct-control policy, using a differentiable kinematic model to let trajectory-level objectives directly guide control learning. Across three simulated obstacle densities, its root-mean-square linear jerk is 14.1%–31.2% lower than the best comparison method on this metric, it achieves the highest tracking success rate, and its per-cycle computation time is 4.25–5.63 ms.
This paper studies line-of-sight pointing control for a two-axis current-command gimbal under mismatched mechanical disturbances. It proposes a quantile disturbance predictor and a tube-inspired model predictive controller that reserves command headroom according to prediction uncertainty. On a two-axis embedded platform, the proposed configuration achieved a 2D pointing RMS error of 3.94 ± 0.10 mrad, with no command-saturation samples within the evaluation window; solve times and cycle overrun records are reported separately for a 100 Hz nominal implementation.
The paper proposes a disturbance-separated model predictive control (MPC) scheme for constrained pointing control of airborne electro-optical gimbals under carrier vibration, input delay, and actuator constraints. The scheme uses an adaptive periodic observer combined with linear-regression frequency estimation to predict multi-harmonic vibrations, and a linear observer plus a bounded LSTM network to estimate aperiodic disturbances and their changes over the input delay; periodic predictions are embedded in the MPC model, while aperiodic compensation is applied outside the optimizer.
The authors propose ISB-NMPC, which couples spherical-bearing prediction with nonlinear quadrotor dynamics to jointly optimize target-approach motion and target visibility. Across 30 paired simulation trials, the method achieved an interception success rate of 86.7%, the highest among the compared controllers, with a mean per-trial target-visible time of 99.11%. In outdoor flights, the interceptor acquired a 30 cm target at approximately 150 m, reached 17 m/s in long-range flight, and flew at approximately 14 m/s while pursuing a maneuvering target towed by a quadrotor. Both software and hardware are open source.
To address the problem that short-term out-of-distribution residuals can corrupt the learned disturbance model during quadrotor trajectory tracking, DGC improves the online update step size while retaining the random Fourier feature model, the offline BPTT policy, and least-squares adaptation (DAPL). It distinguishes adaptation states by comparing long- and short-term gradient memories and uses a confidence map to suppress unreliable updates.
The authors propose ThinkLocoManip, a whole-body control agent framework for legged loco-manipulation that uses an interleaved reasoning-and-acting paradigm to handle long-horizon tasks in large-scale environments. The framework orchestrates perception, navigation, manipulation, and locomotion as composable tools, and it designs cross-domain coordination skills and a keyframe thinking mechanism to support error recovery. In real-robot experiments, its success rate was 77%, more than double the 37.1% of the strongest plan-then-act baseline.
The study proposes HumanoidNDM, an online fine-tuning framework for humanoid motion tracking driven by a neural dynamics model trained on real robots, aimed at closing the dynamics gap between training and deployment. The method includes low-frequency reference perturbation data collection with an adaptive intensity mechanism, training a dynamics model based on proprioceptive history in Euclidean space combined with differentiable forward kinematics, and directly fine-tuning the policy through backpropagated model gradients.