Distribution Transformer Condition Monitoring based on Edge Intelligence for Industrial IoT

Adoption of IoT in industrial applications results in huge volumes of data to be processed. By leveraging edge computing and agent based system architecture, autonomous decisions can be made at edge with locally available data without relying on the cloud. An important aspect to consider while designing smart edge systems is the architecture that enables local intelligence and real-time analytics. This paper proposes an architectural approach that combines the key aspects of edge computing and intelligent agents and presents experiment results using a Proof of Concept (PoC) on condition monitoring of distribution transformers in an industrial setting.