Abstract: A method and a network node for applying machine learning for training a communication policy controlling radio resources for communication of messages between the network node and a control node operating a remotely controlled device is provided. The network node obtains said messages during one or more communication phases communicated when an initial first communication policy is applied for controlling a Quality of Service, QoS, mode. The network node trains a machine learning model based on said messages and the first communication policy. The network node produces a second communication policy including at least one adjusted QoS mode for at least one communication phase. The network node determines a performance score for the second communication policy in the communication phase(s) based on the radio resources used when communicating using the second communication policy.