Data Layer

Gathers energy consumption data, in high frequency, and from various legacy systems that are already deployed on a site.

Back End

The backend a) collects all the available data from the different sources and stores them, b) is responsible for the data analysis using AI components for the Energy Disaggregation, Device Health monitoring and the Decision Support expert system.

Front End

CLARION offers a user-friendly interface with modular dashboards, smart alerts and visualization components to present the results of the analysis and allow easier understanding and interaction of the available data.


The main purpose of the clarion experiment is a) to test the effectiveness of the system towards the reduction in total energy consumption and b) to test the effectiveness of the system towards efficient maintenance planning. The experiment also includes the evaluation of the algorithms and techniques used, and the fine-tuning of the algorithms. The experiment aims to enhance, integrate and cloudify existing energy disaggregation, device health monitoring and DSS tools to provide a market-oriented solution for remote device monitoring and fault detection, and energy efficiency. High frequency energy meters will be deployed to assess the minimum sampling frequency that will allow the disaggregation and device health modules to provide solid outputs. Each partner in the project will have a target role and develop specific components in order to successfully develop and integrate the CLARION system. The actual experiment will take place at the premises of the ELSAP pomace treatment factory in the region of Nafplion in Greece.

IoT dataset

A high-frequency industrial dataset for electric motors, to be used for the benchmarking of industrial energy disaggregation algorithms – Though there are available datasets with domestic and commercial data, there is lack of industrial data.

Algorithms and different techniques evaluation

Validation of the device health monitoring techniques towards predictive maintenance of EMDS using high-frequency electricity load sampling - Evaluation of algorithms and methods that have been used against the proposed ML/DL approach.

High-frequency meter

A high-frequency sampling device (meter) for electric motor systems, tested at industrial scale and industrial conditions - To be used for advanced analytics within the current framework or as input to other tools.

ML/AI Algorithm Framework

An analysis framework using ML/AI algorithms to deliver advanced diagnostics for EMDS - To improve the motor operational efficiency and support predictive maintenance features.

Clarion Project

is a co-funded Application Experiment under the CloudiFacturing Open Call 2 (CloudiFacturing-2)

More Info