

This paper explores the path of industrial product quality and safety supervision from “point” to “surface,” taking the problematic products and enterprises discovered in supervision as the entry point to identify associated and similar risks. The study excavates key factors influencing quality and safety from the perspectives of supply chains and corporate investment and operation, establishes a quality risk relationship transmission network model (quality relationship network model), and uses big data, artificial intelligence technologies (such as community detection algorithms) to estimate the probability distribution of hidden quality states through limited risk point clues. This realizes the measurement of quality risk correlation between industrial products and enterprises, clarifies the risk association degree and impact degree between identified risk points and industrial products/enterprises to be discovered, and further analyzes risk transmission paths. Different from traditional quality supervision, which discovers isolated quality risk points through methods such as quality sampling and public opinion risk monitoring, supervision based on the quality relationship network constructs networks from two dimensions: supply chains and investment operations. Through this network, it identifies and locates enterprises and products with highly correlated quality risks, achieves cross-product type, cross-enterprise type, and cross-regional quality risk identification, and promotes the leap of quality supervision from “point” to “surface,” enhancing supervision coverage. Additionally, the paper proposes suggestions such as accelerating the interconnection and interoperability of supervision data and improving the efficiency of quality and safety risk management through AI research and development based on the research content.