The CTC Technology Centre is working on the creation of a predictive maintenance system based on virtual models. The project, called MOVIDICT (Virtual Models Applied to Predictive Maintenance Systems for Industrial Components), aims to develop and experimentally validate this new system applied to rotating mechanical components, such as those found in pumps or motors, to predict the behaviour of the elements once they enter into service.
Numerous predictive maintenance techniques are available in the industry. All these techniques use a monitoring system and parameters indicating the performance of the machine or component. Depending on the predictive maintenance goal, there are two approaches: fault diagnosis and fault prognosis.
Most techniques are condition or status monitoring and apply common fault diagnosis methods to detect errors and identify their severity. Prognostic techniques allow predicting when a failure will occur based on the current and past condition of the machine. These techniques are based on methodologies from different technologies, such as numerical models, machine learning algorithms, expert systems, or a combination of technologies.
The main innovative challenge of the MOVIDIC project lies in the mission to combine advanced modelling and simulation technology with predictive maintenance techniques. The objective is to develop a predictive maintenance solution based on virtual models of the components. This solution allows characterising and predicting the defects that may occur during the element’s life cycle and its degradation.
Predicting and preventing failures in a component or machine ensures quick and inexpensive interventions, which can save up to 40% in maintenance costs and generates greater efficiency for the company involved and optimisation of the work of the operators.
CTC has a strategic position in advanced numerical modelling and simulation. Specifically, the only technology centre in Cantabria was distinguished by the Cervera Network of Excellence as a centre of excellence in the development and transfer of knowledge on digital twins in the industrial field.
Likewise, through this project, CTC seeks to advance in the development and validation of advanced Machine Learning and numerical simulation techniques applied to complex systems and focused on maintaining industrial assets.
The MOVIDICT project has been funded through the call for R&D projects in the ICT 2020 sector, promoted by SODERCAN.