Abstract: With the acceleration of data integration in various industries under the background of industrial Internet, the continuous expansion of the coverage of railway information system business, and the increasing demand for external cross domain business data circulation, supply chain data security has gradually become a concern of railway digital security systems. Privacy computing, as an important technology to ensure the secure flow of data, can effectively solve the problem of data flow security between railway and supply chain enterprises with its transaction advantage of "data available but invisible". The article conducts research and analysis on secure multi-party computing, federated learning, trusted execution environment and other technologies related to privacy computing. Referring to relevant standards of privacy computing, it proposes a railway privacy computing technology architecture and explores the application of privacy computing technology in the railway industry supply chain. The aim is to provide a security barrier for railway enterprises and external business stakeholders to carry out data sharing applications, promote cross domain secure and efficient flow of railway data, and further leverage the value doubling and spillover effects of the data.