Project

MODA: Model-Bases Data Analytics

Acronym
MODA
Duration
01 January 2018 → 30 September 2020
Funding
Regional and community funding: various
Research disciplines
  • Engineering and technology
    • Electronics
Keywords
data analytics
 
Project description

 

The aim of this project is to develop methods, tools and guidelines (the set of instruments) to enable building a 'digital twin' of a cyber-physical system. The digital twin models how KPIs such as output quality, energy efficiency, production efficiency or availability relate to the settings, context and process measurements. These instruments will be developed for machines that carry out thermal and / or mechanical processes and must at least offer the following options:

  • Data processing of at least 1010 scalar data points consisting of: process measurements, product measurements, measurements of functional product performance, machine settings and context measurements.
  • The physical modelling of a chain of 10 connected machines, inclusive
    • Subsystems of subsystems,
    • interconnected 1D models representing electromechanical and thermodynamic energy flows and the thermo-mechanical material transformation,
    • the ability to introduce hidden variables and parameters,
    • The ability to identify normal and abnormal operation.
  • Support with decisions to
    • Improve machine performance by defining at least three settings when affected by at least two context variables,
    • Define the optimal decisions for events with a probability of less than 10-4 per time step of data.

With this toolkit, the industry will be able to develop digital twins that enable companies to improve the performance of their machines by leveraging their data and core know-how. They will be able to optimize settings for different operational contexts, define corrective actions for drift, errors and anomalies, make extrapolations from one machine to another, and improve the observability of important product, process and machine parameters. This will increase the global operational efficiency of the machines.