As software as a service MITEE will build digital twins of industrial assets and processes, such as anomaly detection models for predictive maintenance and time-dependent regression models for optimizing energy consumption. To ensure accuracy and robustness MITEE will mix deterministic and machine learning models.
Ideal users are facility technicians and managers. In custom dashboards users will track key performance metrics in real time and interact with the developed models. Special plots will help users understand the impact of input variables on the output of these models.
Accuracy and a user-centered approach will set MITEE apart.
Usage of Standards for data interoperability:
- Support of protocols gRPC, RDF, HTTP+JSON
- Design of REST APIs with standard OpenAPI specification
- Definition of digital twins using ontologies such as FIWARE’s Smart Data Models and Azure DTDL