Last news about IDEKO.
The Basque research centre IDEKO is participating in the COGNIPLANT project, which will develop a cognitive monitoring and control platform for the process industry.
The initiative with a budget of 8.5 million euros, will apply advanced data analysis to extract information about the processes and their effect on the performance of production plants.
The process industry, which encompasses activities such as refining, water treatment, chemistry, metallurgy and the manufacture of building materials industry, still presents a wide potential in the implementation of digital technologies.
The COGNIPLANT project was set up with a view to promoting intelligent management of factories in these sectors, an initiative in which the research centre specialising in Advanced Manufacturing, IDEKO, is participating.
The project features a new approach to carry out advanced digitisation of the process industry, through the development of a monitoring and cognitive control platform.
"Digitisation of the process industry still has untapped potential. It is estimated that 70% of the data collected is not used and that many still continue to rely on traditional, poorly optimised procedures. The shortage of raw materials, the rise of energy costs and environmental restrictions require this sector to improve its performance and flexibility", says Iker Mancisidor, the head of the initiative at IDEKO.
In this scenario, COGNIPLANT is to develop an advanced hierarchical monitoring and supervisory control system to obtain a global vision, both of the production performance of the plants, and of energy and resource consumption.
“A novel vision will be established for data analysis and monitoring with the latest developments in advanced analytics and cognitive reasoning, together with a disruptive use of the Digital Twin concept, to improve the operational performance of production plants by up to 68% in the real-time control of the production environment, 65% in quality control of final products and 70% in the response time to unforeseen or uncontrolled incidents," adds Iker.
To achieve this, advanced data analysis will be used to extract information from processes and the effect on the overall performance of production plants, and an innovative digital twin model of factories will be developed.
"The conclusions drawn from the
advanced data analysis will allow us to design and simulate optimal operational
plans on the digital twin models that will be combined with a reactive
real-time planning tool. The conclusions of these operations will be compared
with the patterns extracted from the advanced data analyses", adds the
Dynamics and Control researcher at IDEKO.
IDEKO's input
In this project, IDEKO will lead the digitisation tasks, focusing on designing an advanced and effective strategy for digitisation, and will contribute to the development of different solutions for vibratory problems with the aim of improving the performance and flexibility of the participants in the initiative.
"We will implement our proven expertise in dynamic machine measurements to optimise plant monitoring systems. We will offer the possibility of using various types of commercial sensors for different applications and we will take care of the development of virtual sensors, adapted to the needs and circumstances of each particular plant," says Iker.
The COGNIPLANT concept will be implemented by four end users from different SPIRE industries (Sustainable Process Industry through Resource and Energy Efficiency). Eventually, the resulting solution will be tested in a chemical plant in Austria, an aluminium refinery in Ireland, a building material factory in Italy and a steel plant in Spain.
Three levels of action
The architecture of the COGNIPLANT will consist of the interaction of three general levels that will be customised according to the needs of each specific sector. These levels will follow a data-driven methodology that consists of collecting information from the plant, and using this to make decisions based on the outcomes of the analyses and simulations.
The digitisation layer (Co-Digitise) will be the closest to the production plants and is responsible for connecting the real world (equipment, sensors, etc.) to the virtual world. On the one hand, it will contain the data acquisition structure needed to collect the required data from the production plant. And on the other hand, it will contain the data virtualisation layer to structure the information in a common language for future analysis.
The analysis layer (Co-Analyse) will develop a global methodology for the processing of the data to enable secure and reliable real-time decision making.
Finally, the decision-making layer (Co-Decide) will focus on creating the interface for plant managers. The core of this layer will be divided between a virtual twin of the plant to predict the impact of prescriptive decisions and a reactive scheduling system capable of dynamically accommodating operational plans to the current state of the plants.
"This is an ecosystem capable of detecting deviations, identifying inefficient production processes and providing information on process performance, maintenance or product quality," adds the researcher.
In addition to IDEKO, the consortium of COGNIPLANT includes Ibermática, Ingeteam Power Technology, Savvy Data Systems, Mr Nec, Logpickr, TU Muenchen, SCCH, Hermes Schleifmittel, Fornaci Calce Grigolin, Aughinish Alumina, Acería de Álava, Stam and Core Innovation.
The initiative, part of the H2020 Programme, is funded by the European Commission and has a budget of over 8.5 million euros.
.jpg)
