Reinforcement learning (RL) has evolved beyond monolithic training, yet existing frameworks remain limited to single algorithms or simple offline-to-online transitions. We present multi-phase RL, a framework that orchestrates multiple learning phases for continual policy improvement. It enables efficient fine-tuning of pretrained policies with new data and smooth adaptation from simulation to real-world environments. To support this paradigm, we introduce RL-Studio, a platform that addresses key implementation barriers, including neural architecture mismatches, parameter transfer complexities, and experiment management overhead. It provides phase orchestration, transition-point monitoring, and full experiment lineage tracking. We demonstrate the effectiveness of multi-phase RL through representative scenarios and highlight RL-Studio’s capabilities.

A robotic work cell with multiple robots can accomplish complex tasks that a single robot cannot manage. Robots in the work cell cooperate with each other, or compete for shared motion path, resources and workspace. So coordinating multiple robots is essential for designing a robotic work cell. We present a logic model of a robotic work cell to build controllers that meet required behavior, and construct a performance model to measure the performance of controllers. We verify the presented model with an experimental multi-robot work cell.