EdgER: Entity Resolution at the Edge for Next Generation Web Systems

Thanks to the advances of emerging technologies like Edge and Cloud Computing and microservice development, web architectures are evolving to support Big Data platforms fully. The decentralization of the traditional cloud-centric approach pushes microservices towards the edge of the Internet when constraints exist regarding response time, security, and proximity services availability. This work proposes an approach for detecting anomalies at the edge called EdgER based on entity resolution techniques. Such an approach can be adopted for any application providing proximity-added-value service, even in near real-time. In this sense, a case study is provided in the photovoltaics domain. By implementing EdgER’s building blocks at the edge, it is possible to reconcile the received IoT data streams early against a predefined data model and identify near real-time anomalies in energy production and/or in smart photovoltaic panel operations. In general, EdgeER can also significantly contribute to improving the quality of the data managed by the overall Big Data platform and facilitating the implementation of privacy-preserving proximity-added-value services.