Medical

devices

Process monitoring and

quality control automation

One of the world’s leading producers of medical devices works on innovation, to guarantee increasing precision, performance and trustability to its products for the healthcare services. Thus, the company engaged in an Industry 5.0 project focused on process monitoring and quality control automation of their medical devices production to optimise its manufacturing process with the aim to increase its quality control system. Our experience, summerized in FactorAI (our trustable solution in the manufacturing field), helped developing an effective custom software solution.

Our Solution

Challenge

The product is the result of complex technological solutions, based on a variety of disciplines and materials. The search for solutions that meet specific performances and financial requirements is the most delicate aspect in its development.

Our challenge is to optimise production through automation and robotics in order to guarantee to the client increased industrial competitiveness in their field by:

increasing efficiency of the production process by increasing yield and reducing waste;
improving working conditions in ‘labour-intensive’ and ‘operator-oriented’ industrial processes;
increasing product quality and reliability.

Data Understanding

This medical device’s production process generates many different types of data that are stored in various disconnected databases. Hence it was necessary to start with the collection and systematisation of all data sources. Therefore the first project step the organisation engaged with, was the development of a major infrastructure, capable of integrating all databases and all quality tests of each production phase, following with a massive pre-processing and normalising phase.

Statistical Process Control

On this data collection platform we’ve developed a whole series of indicators and KPIs, both univariate and multivariate, allowing us to effectively monitor all production phases.

Above all, we transformed the results of the various quality tests from a simple go-ahead for production, to indications for optimising production itself. Hence the distribution of quality tests’ values was analysed to support anomaly pattern recognition and the identification of the issue.

Anomaly pattern recognition

The quality control system turns around a 2D Digital Twin of the product, developed by us to visualise the mapping of faults and anomalies per production batch, extracted by production signals analysis. This tool supports operators with practical suggestions of immediate intervention to avoid anomalies and optimise the production process.

The whole system was integrated by a general dashboard, including the overall vision of the whole production trends and performances, thanks to the development of performance KPI that enables the management of the company to identify areas of improvement and development in the production process.

Results

The complete process monitoring and quality control system developed for the production of these medical devices, provided production with a very punctual process control system, allowing: