How AI is Checking What We Eat

With the help of automation and AI, manufacturers now have a window of opportunity to inject more intelligence at virtually every level of the food supply.

Stock Grocery Delivery
Getty Images

In the past, the food industry has come under pressure. The pandemic has not only shut down our favorite restaurants, but it has also caused considerable stress on food demand, production, processing and distribution. This happened by way of the synergistic confluence of several factors, including changing consumer demands, restrictions in movement of workers, limitations on the workforce (number of employees and their distancing) able to work on the production line, and in the worse cases, factory closures.

The resilience of U.S. supply chain in general, and food availability in particular, has already emerged at the forefront of 2021 White House priorities. To make U.S. manufacturers and supply chain more robust, automation and artificial intelligence (AI) are key tools. From facilitating an agile procurement strategy to ensuring last-mile delivery and offering personalized solutions to upstream and downstream supply chain stakeholders, among others, AI and automation have the potential to significantly improve supply chain operations.  

In the food industry, AI can be applied as early as prepping farmland to help create optimal growing conditions to produce better food, as is the case of AI being used to monitor the level of salinity, light, heat and water stress on plants. The systematic adoption of AI and automation in later stages can increase productivity even more efficiently than human workers, as well as monitor the unsafe or low-quality products in the food supply chain and help track a product’s origin from production to end consumer.

Another area where AI can have a very important impact is aiding human workers in the process of visually inspecting production. A study released by the UK government has highlighted how human error on the production line - mostly due to poor training and standardization among the workforce – is at the root of the 10.2 million tons of food wasted annually in that country alone. It is no surprise that global manufacturers are now quickly turning to technology to fix the issue and defect-proof their production line for a future where the human workforce might be even less available than today.

The good news is that AI is ready for the challenge. Today, visual AI software can be easily implemented and integrated in existing infrastructure (inexpensive cameras, industrial PCs) to, as an example, inspect raw materials such as meat for foreign contaminants (e.g. plastic particles), a process not performed today in many instances due to the sheer volume and speed of food being processed. At a later stage, AI can be used to inspect case packing with overwrap correctness of canned foods to ensure the palletizing process is not interrupted by poor overwrapping, and therefore maintain production goals.

Another food production process only subject to spot checking is in baked goods. Human oversight is currently insufficient when it comes to ensuring quality and the even distribution of toppings (sesame seeds, etc.). And, all the way down the production line, visual AI can again be used to compare barcodes of products – such as powdered milk - with their packaging barcodes and verify that the correct expiration date is assigned to the right product.

The past year has seen factory closures, shifts in demand, high absenteeism among workers, price increases and rationing of supplies to avoid stock-outs. As a result, it has become evident that our highly specialized and efficient food processing systems were not flexible enough to quickly adjust to our new reality.

With the help of automation and AI, manufacturers now have a window of opportunity to – quite literally – inject more intelligence at virtually every level of the food supply and ensure that this key pillar of the economy is ready to handle whatever comes its way.