Food Inspections and Blockchain
Many of us do not consider the processes and regulations surrounding the manufacturing of the foods we consume. Agencies such as the Canadian Food Inspection Agency (CFIA) are in place to help regulate this space here in Canada. These agencies have nothing more than the best of intentions to keep consumers safe. They offer guidelines and literal inspections at factories to ensure adherence to the rules, among a variety of other formalities in their efforts to ensure that our food is safeguarded.
As many households become more conscious of the ingredients in their food and it’s nutritional content, they become more reliant on the only thing that gives them insight into the product, the nutritional and ingredient label on the product itself. It may then concern you to know that in most ingredients or nutritional listings that make up this label, it is deemed acceptable by the regulators to be within +/- 20% of the stated quantity. That is to say that if you claim there is 1000 of something, there could be anywhere from 800-1200 of that component.
This is nothing short of alarming, especially when considering the frequency of which these tests are completed. They are limited and infrequent, and most certainly do not account for dynamic change in process, materials, and many more variables involved in day to day production.
An even larger issue can be seen in the following note, present inside the inspection instructions used during ingredient confirmations.
"If possible, observe a batch of product as it is being prepared to confirm the actual amount of each ingredient added. If this is not possible, use the company's make sheet (that is, the form employees complete when they prepare a batch of product) from a previous production run or the batch formula used by employees on the floor."
Industrial automation and manufacturing is a well structured process, but it is definitely not perfect. Are we really relying on the manufacturer to show their data of the previous process to ensure regulations are being met?
There are at least two flaws in the above assumptions: you are trusting the manufacturer to provide legitimate data (it's not always in their best interest to do so), and you are trusting the sensors and automated equipment (or human measurements) to produce the correct data.
With the integration of novel systems using verifiable redundant sensors, certifiable blockchain gateways, confidence fabrics, contextualized production data, as well as a scalable and adaptable blockchain, a plant of the future could allow for "self certification". This will allow manufacturers to capture the data they already plan on capturing in their Industry 4.0 evolution, but gain measurable value from adding this additional process.
This idea is being explored in great detail, and concepts are being designed around the creation of a contextualized data framework that will easily package data into verifiable data stacks, which can be used to adhere to all regulations in place now, and in the future.
This is beneficial for the manufacturer as they are able to self certify, requiring very limited regulatory oversight. They can track recalls in live time, and track the history easily through the hierarchical contextualized data framework. There is also an opportunity to pay less insurance due to the higher and accurate quality, and gain a competitive advantage when compared to their competition. The timing of this new integration is also extremely advantageous for some, as they are still in their Industry 4.0 transition, and now is an opportune time to add these valuable additions.
This will also benefit the regulator as they can ensure the products are being manufactured to their standards, in live time. The resolution gained will be exponentially greater, as yearly inspections will now be occurring everyday (or week) as necessary. Through the use of blockchain and verifiable sensor / processor data, the regulators can be assured in the accuracy of the data. You will also see, if you choose to read some of the links within this post, that regulators are usually striving for the most accurate of readings they can possibly get. They do however, need to settle for "good manufacturing processes" and higher than ideal tolerances because of the manufacturing process and the inability to do it as accurately as arguably possible.
The manufacturing space is fantastic and has led to some spectacular increases in efficiency and throughput, but it has classically (but not always) been motivated more by production output, and regulatory compliance was a thorn in the side that was seen to slow down innovation and advancement in the space of production.
The above concept not only benefits the manufacturer and the regulator, but it also greatly benefits the consumer. Foremost, it gives confidence in the label that leads to purchasing decisions. It also creates a trust that the ingredients are accurate, and the nutritional content is correct. If you can agree that food is our fuel, then the food label is the insight into the type and quality of the fuel the food possesses. Wouldn't it be great to know that the information is accurate?
This concept also has great benefit to the supply chain, and this level of information (from the manufacturing plant level) adds extremely improved resolution into the data and ensures an accurate product for the rest of the value chain.
We hope this post is insightful. More will be coming in the future as the space of blockchain and Industry 4.0 is moving quickly and is ripe for innovation! If you agree with these ideas or would like to discuss them further, feel free to reach out!