ALT Modular reconfigurable robotic systems provide a scalable solution for cooperative surface operations in future lunar missions. However, cooperative cargo transportation remains challenging due to morphology-dependent topology changes, strong payload-induced coupling, long-horizon decision making, and safety constraints. This paper proposes a phase-decomposed reinforcement learning framework for cooperative cargo transport with distributed robotic units. The task is decomposed into lifting, transportation, and placement, each optimized with a dedicated joint-state policy capturing inter-agent coupling. Centralized training promotes stable convergence, while deployment uses onboard proprioception for control and OptiTrack motion capture for ground-truth evaluation and post-processed metrics. A deterministic phase controller expressed in Markov state representation regulates transitions between stages, and a failure-sensitive synchronization mechanism ensures coordinated progression and

ALT Rolling motion planning is challenging because rolling contact imposes nonholonomic constraints and the configuration evolves on a curved manifold. The problem becomes substantially harder in multi-contact settings, where multiple bodies roll without slip and the contact states are coupled. This paper presents a new framework for multi-contact path planning in spherical rolling robotics under no-slip constraints. We first derive a compact kinematic model for multi-sphere rolling using Montana's contact-coordinate formulation, where each contact is represented by a stacked five-state vector. Building on this model, we construct a Voronoi-based roadmap directly on the spherical contact manifold, incorporating spherical-cap obstacles and mutual-exclusion regions via on-manifold collision checking, and refine discrete graph paths using manifold-consistent log-exp smoothing. The resulting smoothed surface paths are then lifted to admissible multi-contact rolling motions through the derived

ALT Keypose-based manipulation decomposes tasks into critical waypoints to simplify policy learning for long-horizon tasks, but existing approaches rely on task-specific heuristics or manual annotation to extract keyposes from demonstrations. We present an automatic trajectory labelling pipeline for grasp-related tasks. This pipeline combines vision-language models (VLMs) for semantic event detection with classical trajectory analysis for precise temporal alignment, requiring VLM inference only on one single demo among repeating ones per task. Using the labelled data, we train a keypose-guided Diffusion Policy (DP) that exploits keypose conditioning to intervene demonstration distributions. We explore the possibility to apply this property for cross-embodiment transfer: candidate keyposes are sampled and filtered via a reachability map, steering the policy toward kinematically feasible keyposes for the target robot. As a preliminary feasibility study, experiments on two robomimic tasks sho