CEBEDEAU • Data processing

In brief

Monitoring water treatment facilities—whether probes, sampling, or environmental quality monitoring—generates large volumes of data. The same is true for biodiversity monitoring and molecular biology analyses.

Our research center helps you process and interpret this data. We analyze it, define your indicators, and transform it into predictive tools for simpler, more accurate, and more efficient water management.

Our approach

Creating experimental design

Every measurement has a cost. Obtaining relevant information for the minimum investment is the raison d'être of experimental designs. How many probes should be installed? Where? What are the critical parameters? Should the measurements be repeated? We answer these questions by applying the principles of DOE (Design of Experiment).

Mathematical modelling

Modeling treatment processes is a powerful tool for understanding changes, sizing new facilities, and optimizing costs. It also allows us to evaluate different scenarios and optimize plant operations, particularly in terms of energy consumption.

We use software such as WEST® and GPSX to simulate processes and provide key information. Our experts can code processes and integrate innovative technologies into the software.

Use of Artificial Intelligence

We develop AI-based predictive tools for real-time regulation and control of large treatment facilities with the aim of reducing GHG emissions and energy bills.

Translating data into dashboards

We configure customized dashboards that highlight the essential parameters of your water facilities with dynamic graphical visualizations. This allows everyone in your organization to quickly view key performance indicators (KPIs), track changes, and make the right decisions.

We use Power BI in particular, but also languages such as R, DAX, SQL, QGIS, and Python.

Statistical data processing

Extracting relevant information from large volumes of raw data is complex. Our experts perform statistical analyses on public or customer-provided datasets, taking into account their needs to provide accurate, simple, and targeted answers.