A task-space receding-horizon controller that uses a short contact-consistent rollout to generate a terminal kinematic reference satisfying internal non-penetration constraints, then computes only the first input of a smooth minimum-acceleration transition toward that reference.
Abstract
Real-time collision avoidance for robotic manipulators requires fast reactions to unexpected obstacle motion and lookahead to avoid becoming trapped by near-future constraints. Full model predictive control can provide this foresight, but its online cost may grow quickly with horizon length, model fidelity, and the number of active geometric constraints. Conversely, horizon-free reactive methods are computationally efficient but can be short-sighted in dynamic clutter. We present a task-space receding-horizon controller that uses a short contact-consistent rollout to generate a terminal kinematic reference satisfying internal non-penetration constraints, then computes only the first input of a smooth minimum-acceleration transition toward that reference. Starting from a closed-loop inverse-kinematics regulation law, the rollout is performed with an iterative dynamics solver operating on inflated convex robot and obstacle geometries, so that robot-obstacle contacts, dynamic obstacle motion, and self-collisions can shape the terminal reference without requiring full constrained trajectory optimization. We analyze the contact-inactive closed loop and show local exponential task-space regulation under standard regularity assumptions. For contacts activated inside the rollout, we characterize the corresponding discrete updates and bound the effect of moving obstacles on regular operating sets. Simulations on a 40-DOF multi-chain system show that intermediate horizons balance anticipation, responsiveness, and computational cost. Hardware experiments on a 6-DOF platform demonstrate consistent sim-to-real behavior without accurate inertial parameter estimation, and comparisons against dynamic optimization fabrics and model predictive control (MPC) baselines show improved success rates in dynamic clutter while preserving solve times compatible with real-time execution in the tested regimes.
The Safe-Koopman Framework is introduced, an operator-theoretic motion planning method that generalizes obstacle-free demonstrations to planar planning tasks with obstacles and avoids collisions observed in the unconstrained Koopman baseline.
Xucheng Liu, Ruiqi Ke, Yandong Wang et al.· IEEE Transactions on Neural...· 0 citations
This paper presents a dual-layered computational framework for the robust trajectory planning and active stabilization of a robotic manipulator transporting a non-fixed payload. The primary challenge addresses the transport of a tray containing multiple objects prone to sliding, exacerbated by significant uncertainties in the system’s dynamic parameters, such as objects’ mass and inertia. The first contribution is an optimal closed-loop sensitivity-based trajectory planning algorithm that generates energy-efficient paths while minimizing the possibility of object sliding. The second contribution is an active sliding control strategy based on Nonlinear Model Predictive Control (NMPC). This algorithm dynamically adjusts the orientation of the tray, mounted to the robot’s end effector, usefully exploiting a dynamic model including inertial forces and gravity to move the object to given positions. Simulation and experimental results demonstrate that the integrated approach allows the robot to set the objects in desired positions with an average steady-state error of 6.3×10−3 m across a set of ten experiments, while limiting the sliding to less than 3% of the tray dimensions during successive transportation in face of a 10% uncertainty about objects’ masses. The synergy between the uncertainty-aware planner and the NMPC controller supports robust tray-transport tasks in unstructured environments where precise dynamic modeling is unavailable.
Sara Federico, Ciro Natale, Fabio Ruggiero et al.· Machines· 0 citations
Autonomous trajectory planning for unmanned surface vehicles (USVs) in complex environments must ensure collision safety, kinematic feasibility, and smooth motion. This paper presents an optimal-control trajectory optimization method implemented in CasADi and solved with the IPOPT interior-point algorithm. The USV is represented by a simplified planar kinematic model with position, heading, and surge speed as states. Longitudinal acceleration and yaw rate are used as the control variables. Obstacle avoidance is modeled through squared-distance inequality constraints for circular obstacles inflated by safety radii. The planning problem is formulated as a finite-horizon nonlinear program with objectives for terminal accuracy, control effort, control smoothness, and path compactness. By optimizing path generation and kinematic feasibility in a single problem, the formulation avoids the decoupling common in staged planning pipelines. Simulations in five scenarios, including single static, multiple static, narrow-corridor, single dynamic, and multi-dynamic obstacle cases, validate the method. Across all scenarios, the optimizer achieved sub-millimeter terminal errors while maintaining positive obstacle clearances. Parameter sensitivity analysis shows a predictable trade-off between safety margin and path length, and ablation experiments quantify each objective term’s contribution. These results indicate that the proposed formulation can generate safe, feasible, and smooth USV trajectories in representative complex environments.
Zhenhao Yan· Advances in Engineering Tech...· 0 citations
This work presents an adaptive formation and collision avoidance framework for differential-drive multi-robot systems. The strategy is based on a deformable virtual structure for planning, control barrier functions for safety, and a decentralized model predictive control to drive robot motion. The virtual structure acts as a centralized planner that generates smooth, collision-aware slot trajectories, while each robot uses a local model predictive control to track its assigned reference under kinematic and wheel-speed constraints. Similarly, control barrier functions modify motion and deformation states to guarantee obstacle evasion without requiring rigid formation constraints. The method is validated experimentally using four differential drive mobile robots navigating multiple obstacles, demonstrating safe deformation, coordinated motion, and tracking despite non-ideal initial poses and continuous reference movement. The results suggest that the proposed navigation and control scheme offers a scalable, adaptive, and robust approach for multi-robot formation control with collision avoidance.
C. Canseco, Herman Castañeda· 2026 IEEE/ASME International...· 0 citations
Trajectory planning is a key enabling technology for UAV swarms operating in complex and obstacle-rich environments. This paper proposes a distributed nonlinear model predictive control (NMPC)-based trajectory planning method for UAV swarms, where terminal target reaching, prescribed formation maintenance, obstacle avoidance, and inter-UAV collision avoidance are incorporated into a unified predictive optimization framework. To reduce the online computational burden, a control parameterization strategy is introduced to describe the control input using M control segments, thereby reducing the dimension of the online decision variables. Furthermore, an exact-penalty-based constraint transcription method is developed to transform the original constrained optimal control problem into a lower-complexity finite-dimensional nonlinear programming problem, while efficiently handling velocity constraints, obstacle avoidance constraints, and inter-UAV collision avoidance constraints. Simulation results for a three-UAV swarm in a cluttered environment demonstrate that the proposed method can generate dynamically feasible and collision-free trajectories, while enabling the swarm to reach the assigned target positions and preserve the desired formation within a certain formation error. Furthermore, real UAV swarm flight experiments were conducted to further validate the practical feasibility and online applicability of the proposed distributed NMPC framework for UAV swarm trajectory planning.
Yingting Cui, Tong-Xin Zeng, Bin Li· Drones· 0 citations
This study proposes a novel real-time trajectory planning framework designed to address the complex navigation challenges of collaborative robots operating in shared workspaces. The developed framework effectively integrates Model Predictive Path Integral (MPPI) control with Artificial Potential Field (APF) mechanisms to ensure robust and adaptive motion generation. To comprehensively evaluate its performance, the approach was validated in diverse environments, including no obstacle, high multi obstacle, and wall configurations, using complex setups of two, three, and four UR5 manipulators. Throughout these extensive tests, the proposed algorithm consistently achieved a 100% success rate in real-time collision avoidance while strictly adhering to the inherent kinematic limits of the robots. Comparative analyses further highlighted the superior efficiency of this method. In the most challenging environment, the proposed framework outperformed a baseline jerk-and-acceleration controller by completing operations 21.3% faster with a 3.8% lower jerk, and significantly surpassed a pure acceleration controller with a 41% faster operation time alongside a 16.7% reduction in jerk.
Nezih Bora Yavas, Zafer Bingul· 2026 IEEE/ASME International...· 0 citations