Reinforcement learning with mone (RLWM) - Pump