c) Integrate planning and control algorithms with differentiable simulation technologies to build a trajectory data collection system that generates high-quality, diverse task data for vision-language-action learning, enabling seamless integration from human instructions to robotic actions in complex mobile manipulation scenarios.