Data Analytics – Leveraging MES Data to reduce CAPAs and Non-Conformances


Our customer, a major MedTech corporation manufacture a very wide range of SKUs rapidly and in very large quantities.  As a result, they need to pick up on defects fast. They’ve a full MES implementation and an excellent process for testing product across a wide range of attributes. However, natural variability in the product means detecting production issues isn’t easy.  And once, a potential issue is detected ME can spend a lot of time working through data sources in order to confirm a theory or build a picture.  The data was there in the MES system but it wasn’t working for them.


Many MES implementations move from a Paper Batch Record process to an Electronic Batch Record (EBR) to reduce the cost of compliance and yield savings in the Batch Review process. They focus on the Operational side too – utilizing the data to help with performance monitoring and Continuous Improvement.

On the other hand, CI projects often focus on increasing yield or OEE, or reducing rejects or the amount of downtime.

What about leveraging the MES data to try and predict issues that may occur and avoid costly Non-Conformances or CAPA’s ?

Our client was capturing critical audit and inspection data relating to Critical to Quality product specifications in their MES system. While this data was a critical element of the Batch Record and was used to disposition product, the customer wanted to investigate if it could be leveraged further. Their goal was to progress their level of Digital Maturity towards Level 4 and ultimately Predictive Plant.


A project team was established and Microsoft Tools were leveraged to perform ETL and Aggregation jobs on the MES Data and then store it in a Data Warehouse. SSIS was used to automatically copy the inspection data across A second set of jobs were run on the Level 0 data to manipulate it into Level 1 and Level 2 states for use by the Business Intelligence Tool (Microsoft Power BI).

Power BI was used to build a set of Control Charts based on Nelson Rules. Color coding highlighted the data points that broke any of the 7 implemented rules.  The charts are used by Manufacturing Engineers to analyse the data and track down the root cause of an issue.

However, the real value of the system is the automation of this process. Rules are checked automatically in Power Query and Quality Engineers are notified via a Power BI alert when rules are broken.  In addition, the rules are can be checked simultaneously on multiple groups of data. For example, data can be grouped by machine, raw material or SKU.  In total, over a hundred different data groups are actively monitored on a daily basis.

Get In Touch with Us

Contact Us today to learn more about how leveraging MES Data can reduce CAPAs and Non-Conformances in your organisation.

To stay up to date with Dataworks Limited news and events,

connect with us via the links below or call us on 051 878 555 today. 

DATAWORKS – Driving Digital Transformation in the worlds’ Leading Life Science Companies