The application of artificial intelligence (AI) solutions can significantly improve the success of new and seasonal F&B products – with sustained mainstream applicability anticipated. Presenters speaking at the Institute of Food Technologists’ (IFT) virtual expo, SHIFT20, showcase how AI can be used to analyze these data points to provide direction for NPD  and go-to-market strategies for product launches. Tracking and predicting purchasing behaviors and customer sentiment can help brands identify common characteristics of products that have succeeded, versus those that have failed, the presenters outline.

“Many food companies are already using AI, for instance – for food sorting, food safety monitoring, forecasting and predicting the need for equipment service and maintenance. However, there is still a tremendous amount of opportunity for the forward-thinking companies. AI technology is a component in the Industry 4.0 vision and journey,” Mikael Bengtsson, Industry & Solution Strategy Director at Infor, a SHIFT20 exhibitor, tells FoodIngredientsFirst.

Speaking at SHIFT20, Infor’s representatives underscore the complexities in F&B manufacturing stretching across the conceptualization of a product to its commercialization. These “moving parts,” as they stress, have traditionally been manual processes. An example of this is the use of excel sheets, or paper and pencil, to calculate a food or beverage’s nutrition facts.

“This comes down to the data that’s contained within your formulas, within your raw materials and how it feeds down into the regulatory space. If you don’t have robust systems in place, then you stand to have a lot of issues along the way. You could mislabel a product and run into trouble,” says Eric Krums, Solution Consultant, PLM for Process Manufacturing at Infor.

Krums traces a timeline of process evolution that sees F&B formulators increasingly relinquishing their conventional methods for AI-backed virtual dashboards, and enhanced virtual toolkits. This, he outlines, is useful when tracking formulas. “Using a robust toolset wher one can see across all the information available is the next step. If I’m using a paper-based system or excel, I’m not able to see all that. We’re cutting down on that chance of having some sort of an error because we’ve miskeyed information,” he explains.

“Ultimately, we can put this data into dashboards that would be used by R&D managers or upper levels of the organization to see exactly what’s going on with all this information. But you can’t get to this step if you don’t transform the front end of your process.”

Savings for manufacturing
As a cited case study within its virtual seminar, Infor worked with supplier Flint Hills to help it consolidate its data processes. The US-based biofuels and ingredients company required a better understanding of asset maintenance patterns and unifying siloed machine learning projects. In addition, the company flagged inconsistent data across site-centric databases and “no means of harmonizing that data.”

“They needed a better understanding of their manufacturing techniques when it came to things like monitoring their assets. What we saw is that they did a tiered approach, wher they started connecting their assets with commodity-based hardware which is very cheap and easy to get- and monitoring their equipment. After that, after they started to collect data and push that into our data lake, they started taking their different machine learning processes and tie that together into one platform,” remarks Heath Branum, Product Manager AI and Machine Learning Solutions at Infor.

Ten million of potential savings of inventory reduction and a further US$50 million are to be put toward transforming its maintenance business. “Now they’re able to see data coming in real time, off their manufacturing floor and from their manufacturing processes. Sometimes they can deploy machine learning models on those machines to actively tell them when maintenance needs to be performed or how long their asset has before it is no longer operational,” notes Branum

Looking ahead, Bengtsson envisions, “One of the most exciting areas will be how output can be improved by optimizing farming conditions. AI will play an important role in analyzing the effects of changing different parameters in agriculture. Furthermore, depending on the access to outside data, competitive and emerging market intelligence could even be leveraged to help gain market leadership.”

AI-driven at SHIFT20
Also exhibiting at SHIFT20, dairy flavors manufacturer Edlong is showcasing how the use of AI can help predict the way consumers will respond to particular sensorial properties of NPD. Leveraging this science, the company has expanded its line of dairy-based natural flavors, marketed as a variety of “authentic-tasting, high-intensity” cheese and cultured profiles for use in sauces, dressings and dips.

Scientific instruments supplier Shimadzu will also be exhibiting at SHIFT20. The company is showcasing its tool, the Nexera UHPLC series, which leverages AI and the Internet of Things (IoT) in help with food lab management. On-offer is a portfolio of solutions, including chromatographs, mass spectrometers, molecular and atomic spectrometers, thermal analyzers, total organic carbon (TOC) analyzers, particle size analyzers and materials testers. Whether testing for pesticides, mycotoxins, toxic metals, additives or packaging, Shimadzu specializes in platforms that help companies meet regulatory requirements.

The virtual showfloor is also hosting Gastrograph AI – an AI platform that helps consumer consumer packaged goods (CPG) explore product concepts, predict consumer preferences and identify preferred flavor profiles. By optimizing development effort, Gastrograph AI shortens time to market and reduces the risk of product failure. 

SGS is also showcasing its AI-driven regulatory platform. The inspection, verification, testing and certification company oversees a global network of ISO 17025 accredited testing laboratories. These facilities offer analyses of F&B products against national and international standards. “With a comprehensive range of independent inspection, testing, training, certification and technical services specific for the food sector, we help companies worldwide to monitor and validate safety, quality, and sustainability,” it outlines.