Big Data and Quality Control

How pairing in-line, non-contact quality control with Big Data provides deeper insights into manufacturing.

Bilal Paracha, Interstates Control Systems Inc.
Bilal Paracha, Interstates Control Systems Inc.

Big Data analytics is a popular phrase within the consumer packaging industry, in particular. With recent developments in non-contact in-line quality control, there is merit in combining these two technologies. The combination can provide further insights into the quality of your product.

Whether it is continuous processing, batch processing or high speed machine control, one of the most critical processes in manufacturing is quality control. Historically, product would be sampled off the line, whether by planned sampling or event driven sampling. These samples would go to a lab, and employees would perform various tests to qualify these samples before they are judged consumer ready. Personnel or systems would save the quality data in a non-digital file and store it away. An alternative scenario would have a separate quality data system for long-term storage.

This process works, but it has a few shortcomings. First, this process has a high operation cost. Secondly, the data saved from the lab quality test is disassociated from the control system; this prevents big data traceability between quality and process run data.

The pairing of in-line, non-contact quality control with Big Data provides a deeper insight into the manufacturing line. It also provides integration and historical connection between quality data and process data. For example, if you have a new product launch and part of the qualification of the new product launch involved startup of a new line (requiring quality, process, and equipment qualification), an integrated approach towards in-line quality data and process data can save time and money by moving away from the old line sampling method. The ROI is there to be had and applied!

There are various mechanisms to approach this integration. A historian server can log data, data can reside into a more common SQL or Oracle database, or the site can use both options. The decision of what data warehouse to choose depends on the specific platform chosen for individual manufacturing needs. It also depends on the analytics software chosen and how easily it integrates with the data warehouse.

As traceability moves from an option to a necessary quality, it only makes sense to use the tools now available. This is especially true when it comes to having the ability to measure and evaluate product quality in real time in a way that maintains ties to the manufacturing control system. Whether through a historian server, database, or both, it is important to decide how Big Data analytics plays a part in manufacturing today.

Bilal Paracha is a business intelligence lead for Interstates Control Systems Inc., a Certified member of the Control System Integrators Association. See Interstates’ profile on CSIA's Industrial Automation Exchange.

More in Data