A unified data security framework for federated prognostics and health management in smart manufacturing

Data security is one of the most important concerns of manufacturing entities. Therefore, enterprises hesitate to share data with third parties for building predictive and prognostic models. In such scenario, building comprehensive models for predicting asset failures is challenging given data from a single enterprise would not include required variety of operation regimes and failure modes. In this article, we present an encrypted and federated framework for training diagnosis and prognosis models that would not require sharing data. The proposed framework guarantees the privacy of the data while generates comprehensive models that leverage the variety of multiple enterprises operations. (C) 2020 Society of Manufacturing Engineers (SME). Published by Elsevier Ltd. All rights reserved.