A Federated Learning-Based Industrial Health Prognostics for Heterogeneous Edge Devices Using Matched Feature Extraction

Data-driven industrial health prognostics require rich training data to develop accurate and reliable predictive models. However, stringent data privacy laws and the abundance of edge industrial data necessitate decentralized data utilization. Thus, the industrial health prognostics field is well suited to significantly benefit from federated learning (FL), a decentralized and privacy-preserving learning technique. However, FL-based health prognostics tasks have hardly been investigated due to the complexities of meaningfully aggregating model parameters trained from heterogeneous data to form a high performing federated model. Specifically, data heterogeneity among edge devices, stemming from dissimilar degradation mechanisms and unequal dataset sizes, poses a critical statistical challenge for developing accurate federated models. We propose a pioneering FL-based health prognostic model with a feature similarity-matched parameter aggregation algorithm to discriminatingly learn from heterogeneous edge data. The algorithm searches across the heterogeneous locally trained models and matches neurons with probabilistically similar feature extraction functions first, before selectively averaging them to form the federated model parameters. As the algorithm only averages similar neurons, as opposed to conventional naive averaging of coordinate-wise neurons, the distinct feature extractors of local models are carried over with less dilution to the resultant federated model. Using both cyclic degradation data of Li-ion batteries and non-cyclic data of turbofan engines, we demonstrate that the proposed method yields accuracy improvements as high as 44.5% and 39.3% for state-of-health estimation and remaining useful life estimation, respectively. Note to Practitioners—Data-driven machine health monitoring enabled by cyber-physical systems allows humans to make timely predictive maintenance decisions for optimizing the life cycle of critical industrial assets. However, the benefits of intelligent health prognostics may not be fully realized in practice because strict data privacy laws restrict the aggregation of decentralized industrial data that are essential for model training. In proposing a privacy-preserving learning technique, we also focused on tackling a common and practical problem of learning from heterogeneous degradation data at the edge. The competitive performance of the proposed similarity-matching-based federated learning algorithm indicates its suitability for modeling heterogeneous, industrial time series data. Therefore, industry practitioners can utilize this algorithm as a staple component of their machine health modeling toolkit in an increasingly privacy-concerned era.