This website uses cookies primarily for visitor analytics. Certain pages will ask you to fill in contact details to receive additional information. On these pages you have the option of having the site log your details for future visits. Indicating you want the site to remember your details will place a cookie on your device. To view our full cookie policy, please click here. You can also view it at any time by going to our Contact Us page.

Eliminate the hidden factors impacting quality

21 September 2020

Laura Stridiron explains the role that digitalisation can play in overall product quality as a competitive advantage for the food and beverage industry. 

Maintaining optimum quality standards is one of the most important, and also difficult, challenges facing food manufacturers.  While quality control has defined processing parameters of variables – including time, temperature and good manufacturing practices (GMPs), there are also strict regulatory requirements and a new trend toward ingredient traceability as a key component to building consumer trust.
 
Adding to the challenge, many factors that impact food quality are hidden. Even with specifications defined within a product formulation, minor variations in processing may still arise and these can steer production off course. Despite traditional means of recording process data, quality compromises might arise from minor, but acceptable discrepancies in raw material properties; variations in procedures; and changing environmental conditions. 

Without understanding the exact variances  – which are not always visible to operators – production cannot easily forecast the outcome or effectively compensate, potentially resulting in an expensive defective batch.
 
Improving confidence
Often it is an organisation’s existing data holds the key to improving its confidence in data. Particularly in the food and beverage industry, mountains of data exist to comply with food safety tracking requirements. However, information such as maintenance work orders and compliance reports typically reside elsewhere in an organisation, potentially separated from related material.  
 
This division of data means past efforts to mine it may not have yielded notable results. As digital transformation efforts increase, some companies are looking to this historical information to give them an insight into how to improve quality moving forwards. 
 
So, what is the first step in utilising past production data to provide insight? With traditional data analytic methods, many businesses have to hire data scientists to perform complex analyses. However, in addition to being unrealistic for most organisations, this approach neglects the other subsets of data in companies’ hands: the ‘hidden factories’ or inherent knowledge of those among their staff who are intimately familiar with the ins and outs of the process. 
 
While they may not have been the creator of the product itself, they will be familiar with how to execute on the product formulation given the surrounding circumstances. Even with sophisticated tools gathering and organising key data, a challenging task remains: Analyse the data to determine how best to adjust the batch process for each variable and accomplish an improved orders outcome. With many data visualisation and analysis techniques, it may still be a daunting challenge for operators, engineers and analysts to sift through data patterns and identify how best to adjust processes. 
 
Extracting meaningful data increases production supply quality. The more quickly that raw data can be turned into actionable insights, the better – especially if it can be done without using new resources. Multivariate analytics software can help solve process and product quality issues, which could likely result in greater customer satisfaction. 

Aspen ProMV can offer a solution designed specifically for use by those most familiar with the process. One international food manufacturer uses it for maintaining quality while increasing supply chain flexibility. Traditionally, this food manufacturer tested incoming raw materials as its primary method for predicting overall product quality. However, this methodology was resulting in unreliable outcomes and an unacceptable level of off-spec product. Having multiple raw material suppliers for each ingredient of a specific product, the company was faced with a dilemma on how to best pre-determine the outcome of final product quality.
 
With the availability of historical data of the product’s raw material lots and variable processing conditions, Aspen ProMV was utilised to develop a data model that correlated these two factors. The data associated with the model revealed two raw materials had no significant impact on final quality while three others did. Therefore, the manufacturer was able to determine where to focus its raw material specification efforts to eliminate future off-spec product and adjust its manufacturing process to improve overall quality. The company was able to scale the same modelling technique across their entire product line which helped avoid raw material combinations that would have led to future poor final quality — ultimately avoiding potential customer satisfaction issues. 

Laura Stridiron is senior product manager at AspenTech.


Contact Details and Archive...

Print this page | E-mail this page