Abstract: The network topology of industrial Internet of Things is complex, including multiple levels and types of network devices. Data needs to be forwarded through multiple nodes during transmission, and each node will have a certain processing delay, resulting in an overall increase in data transmission delay. A low latency transmission method for industrial IoT device data based on 5G communication has been proposed to address this issue. Firstly, a communication transmission delay model for industrial Internet of Things was constructed. This model fully considers the high bandwidth and low latency characteristics of 5G communication, as well as the diversity and distribution of industrial IoT devices. Secondly, calculate the limit value of channel capacity based on Shannon′s formula and use it as a reference for data transmission planning. Finally, a device data adaptive delay fair scheduling transmission method was designed. By dynamically adjusting the transmission priority, important data can pass through each node faster, reducing the waiting time at each node and alleviating the delay caused by complex network topology to a certain extent. Case studies have shown that this method can significantly reduce data transmission latency, improve transmission efficiency, while ensuring fairness and reliability of data transmission, providing strong support for the intelligent and automated production of industrial Internet of Things.