On-site smart maintenance based on collaboration between network large model and AI models

On-site maintenance is the end of telecom operators' network maintenance, and it is an important means of ensuring the quality of telecom network operation and user experience. Therefore, improving the quality and efficiency of on-site maintenance is essential. This paper designs an on-site smart maintenance scheme based on the network large model, according to the actual maintenance processes of telecom operators. It also proposes practical use cases such as maintenance Q&A assistant, on-site troubleshooting and analysis assistant, and on-site maintenance agent. By giving full play to generative AI and orchestrating the ability of small models, the network large model helps maintenance personnel solve practical problems in production. It not only improves the maintenance efficiency but also effectively guarantees the safe operation of the network. While improving customer satisfaction, it can also reduce the maintenance costs of operators to a certain extent.