Prior Policy Guided Dual-Agent Coordinated Manipulation Planning of Spacecraft-Manipulator System

Deep reinforcement learning has the potential to solve attitude control problems more adaptively, precisely, and robustly by handling nonlinear dynamics, uncertainties, and failure cases more effectively than traditional attitude control approaches. We explore reinforcement learning (RL) for attitude control in spacecraft re-entry. An industry-standard proportional-integral-derivative controller with gain scheduling serves as a strong baseline for model-free RL and hybrid controllers that combine these two approaches. We formalize the application in the RL framework to apply continuous, off-policy RL. State-of-the-art RL achieves comparable performance to traditional control approaches in this domain. However, its out-of-distribution generalization is not sufficient. Hence, we use dynamics randomization to introduce challenging task variations during training and enforce generalization in a predefined operational envelope. Finally, we assess the best obtained RL-based controllers with application-specific metrics to show superior performance in comparison to traditional controllers in the operational envelope, that is, hybrid controllers are able to track the angle of attack better and are more robust under variations of mass, inertia tensor, and flap actuator bandwidth.

Planning the motion path for a tightly coupled dual-arm space manipulator under closed-chain constraints is a fundamental yet challenging problem in on-orbit assembly of large-scale space structures. The closed-chain constraints significantly reduce the feasible configuration space, making it difficult for existing planners to efficiently generate collision-free motions, especially in cluttered environments. To address this issue, this paper proposes a task-space constrained bidirectional rapidly-exploring random tree algorithm, termed TCBiRRT. Unlike conventional methods that operate in the high-dimensional configuration space, the proposed approach performs random sampling and node expansion directly in the task space defined by the manipulated object pose. A task-space node expansion strategy is developed to generate candidate object motions, which are then mapped to continuous joint paths using a path inverse kinematics algorithm. The method is further integrated with a bidirectional RRT framework and a regrasp mechanism to efficiently connect two random trees. Extensive simulations are conducted in representative on-orbit assembly scenarios with varying levels of environmental complexity. The results demonstrate that TCBiRRT achieves significantly higher success rates and orders-of-magnitude improvements in planning time compared to state-of-the-art planners. The proposed method provides an efficient and robust solution for motion planning of tightly coupled dual-arm space manipulators.

Affordance-guided imitation learning has shown impressive performance in robotic manipulation tasks by compressing visual perception into task-specific geometric constraints (e.g., fixed contact points). However, the commonly used static affordances can become inconsistent in precision-critical tasks or under object location perturbations, leading to post-contact trajectory drift. To address this issue, we propose AffordTrajDP, a dynamic framework that constructs affordance trajectories via object-centric temporal propagation to guide the progressive manipulation process. Specifically, given an RGB-D observation, our core insight is that a retrieved anchor affordance, which captures the desired contact point between the end-effector and the target object, can be propagated forward via affordance propagation, using the object's SE(3) pose as a natural propagation medium, to yield an affordance trajectory that provides temporally consistent, state-aware guidance throughout execution. AffordTrajDP achieves 70.0% average success rate on ManiSkill3, outperforming strong baselines by up to 17.8%. Real-world experiments on Galaxea A1 and UR7e robotic arms, covering StackCube, PickCup, AdapterInsertion, Ring-on-Peg, Put-in-Bowl, and USB Insertion, further validate robustness under object placement variations and appearance changes, with seen and unseen object instances evaluated on Galaxea A1, and ablations confirm the contribution of each proposed component.