There are many aspects of the food industry that require good data management, whether it’s recipe and formulations development, the management of raw ingredients, manufacturing and process control, or post-market feedback.
The reasons for good data management are diverse — typically drivers are focused on speed to market, the innovation of new food products and regulation. In this article we look at the business problems faced by the food industry when considering regulatory compliance.
The impact of the many and diverse regulations around the globe demand that food companies are able to pinpoint the exact contents of any given food product at any point in time. If we expand this demand, it can become very complicated, very quickly. Consider the use of raw materials — these are obtained from suppliers or manufactured in-house. These raw ingredients can then be used to either make another ingredient for a formulation by combining with other raw materials, in a process called ‘batch in a batch’, or sub-assemblies. Raw materials and ingredients are then combined in a formulation (to given ratios) and combined via a recipe (a process). Once combined in the creation step, a product batch is created and packaged up ready to be delivered to distribution and subsequently customers.
This is a very simplified description, and in the real world the process is much more complicated, but the fundamentals are the same: raw materials > ingredients > product. Now, if we add to this the regulatory requirements of being able to know and track the exact batch of a given raw material that was used to create an exact batch of an ingredient, and in turn, used in an exact batch of product, we can start to see the importance of data management. This is compounded by the requirement to answer two specific questions:
- I have a batch of product — tell me everything that went into this; or
- I have a raw ingredient — tell me all the product batches that contain it
These questions conceptually are not difficult, but executing them and having the right data management capabilities to answer these questions is critically important. The data for every batch of every component needs to be captured and, importantly, connected with the data for other batches. This is called genealogy — how things are related.
If the genealogical relationships are captured in a manner that makes them "searchable," then the two critical questions described above can be answered. Equally, the regulatory implications of having to be able to quickly answer the question of "what is in what?" can also be accomplished. However, the key in this statement is quickly.
Where regulation is concerned, questions need to be answered rapidly — it is no longer acceptable to use a manual paper process and take weeks to get the answers. Regulators, like the FDA, expect answers within hours. The reason why the response times are required to be so short is understandable — contaminated food or noncompliant food entering the food chain can have massive implications and intervention needs to be rapid. This regulatory requirement can only be met with good data management: tracking and linking of all raw materials and ingredients batch information.
But, this is only half the story — the other aspect is knowing what the ingredients have touched and how they have been touched.
Contaminants during the production process can cause a product batch to be unacceptable. So batch information needs to be combined with where it has been, adding another level of complexity. However, similarly to the genealogy tracking and searching problem, what a given batch has touched can also be captured and linked using good data management. With both aspects covered, managed and tracked, regulatory requirements can be met confidently and quickly.
Additional benefits of reduced paper burden and improved access to business critical information are also aspects that make a very compelling business case for introducing and leveraging good data management approaches for food companies.
Importantly, data must be connected. With explicit genealogical relationships recorded alongside all the batch context, regulatory requirements can be met in a timely and robust manner.
From Raw Materials to Product Display
Should a defect be spotted during the manufacturing process of any food products, the leveraging of big data in the supply chain will be essential for the company in question. In November 2016, Sabra voluntarily issued a recall on 54 of its hummus products as Listeria bacteria was found in its manufacturing facility in Richmond, Virginia. Tests on finished products revealed no traces of Listeria bacteria but that did not deter Sabra from recalling products with a "best before" date up to January 23, 2017. Sabra also urged consumers in the U.S. and Canada to discard any products that may have been affected.
The cost of product recalls can be detrimental to any organization — in 2008, following a Salmonella outbreak, the Peanut Corp. of America had to file for Chapter 7 bankruptcy, with the head of the Georgia Peanut Commission stating the company could face losses of $1 billion.
Whether it is understanding the problem to a greater depth or allowing a part-recall of any products in question, good data management extends beyond the realms of compliance and can even be used to protect brand reputation and revenue.
Advanced data management could have also helped minimize the effects of the H-E-B Baby Food 2 pack 4-ounce recall incident. After a consumer found a foreign piece of plastic in a 4-ounce cup of baby food, H-E-B issued a recall of 18 2-ounce cup baby food products, offering all affected customers a full refund. Spotting these defects in the manufacturing process and enacting tailored recalls based from batch number can save companies embarrassment and money, and minimize delays to business operations.
The additional benefits of reduced paper burden and improved access to business-critical information are also aspects that make a very compelling business case for introducing and leveraging good data management approaches for food companies.
The bottom line is simple: data must be connected. With explicit genealogical relationships recorded alongside all the batch context, regulatory requirements can be met in a timely and robust manner.
About the author
Paul Denny-Gouldson heads the overall strategic planning for the various market verticals and scientific domains at IDBS. He obtained his Ph.D. in Computational Biology from Essex University in 1996, and has authored more than 25 scientific papers and book chapters