Contributed talk: Adaptive Online Planning for Lifelong Reinforcement Learning