Abstract: In this work, we investigate the effect of sensor-actuator clock offsets on reinforcement learning (RL) enabled cyber-physical systems. In particular, we consider an off-policy RL algorithm ...