Mentor's Musings on Federated Learning Standardization Imperatives in Leveraging It for Industrial IoT
N. Kishor Narang · IEEE Internet of Things Magazine · 2024
Federated learning is regarded as an effective solution to allow a distributed learning scheme without explicit data sharing and private data leakage, which makes it really very useful for Industrial IoT solutions where internet-connected devices and sensors are spread over diverse geographical and/or industrial settings to collect and exchange data, monitor and control processes, and enable more efficient and effective industrial operations. However, structured system level standardization is crucial to provide a blueprint for data usage and model building across organizations and devices while meeting applicable privacy, security and regulatory requirements. The system standards shall enable defining the contours for structured innovation in this new domain and ensure different FL frameworks and their respective components can inter-work seamlessly to make Federated Learning empowered applications truly efficient and comprehensive.