Japan’s Kanematsu Taps 2nd Nature’s AI Tech to Upcycle Ag Waste Into Functional Foods

Japanese trading giant Kanematsu Corporation has teamed up with US startup 2nd Nature to use AI to explore the potential of its agricultural sidestreams as sources of high-value functional ingredients.

Artificial intelligence (AI) is already turning the food industry on its head, opening up opportunities for product development and reformulation, enabling faster, cheaper production and helping keep up with consumer trends.

One of the most promising areas that food companies are looking to AI for is ingredient discovery. It’s an approach pioneered by Chilean unicorn NotCo, which uses the technology to help the world’s largest food companies identify the most suitable ingredient combinations for new and sustainable products.

US startup 2nd Nature is also leveraging AI for ingredient exploration, but with a specific focus on waste. The Ohio-based firm maps the molecular makeup of agriculture and food processing sidestreams to identify “hidden” small molecules, fibres, peptides, proteins, and enzymes.

“Our tech covers a plethora of crops and industry sidestreams. Most of them are major food crops, and food and beverage processing side streams, such as grains, legumes, fruits, and vegetables (for example, tomatoes), as well as fermented F&B and dairy byproducts,” 2nd Nature co-founder and CEO Effendi Leonard told Green Queen.

These can then be used to produce high-value functional ingredients for the food and beverage, health, personal care, and wellness markets. And that’s exactly what underpins 2nd Nature’s partnership with Kanematsu Corporation, a Japanese trading company that operates across multiple sectors, including food and materials.

2nd Nature’s three-step ingredient discovery process

Through the collaboration, Kanematsu is looking to deepen its understanding of how underutilised byproducts from its existing operations could potentially be utilised more sustainably, instead of being discarded.

The Japanese corporation will leverage 2nd Nature’s AgWaste Portal to assess the potential of its sidestreams as sources of functional ingredients. The platform analyses complex chemical data to identify compounds invisible to existing human-led screening methods.

“Food processing sidestreams are treasure troves full of nutrients and high-value ingredients. For too long, they are viewed as waste,” Leonard said in a statement. “Our AI looks for valuable ingredients within wastestreams.”

Its full suite of services blends AI-powered discovery with hardware to streamline ingredient production. Its first step, titled Discover, uncovers functional ingredient opportunities in sidestreams, which is followed by Develop, which entails creating scalable, cost-effective processes to produce target ingredients.

Finally, the Deploy stage provides manufacturers with a de-risked, resilient supply of upcycled ingredients sourced from their own sidestreams, helping them reduce waste, generate additional income streams, and support food sustainability.

The end-to-end approach means the insights generated during the Discovery phase can lead to the creation of the shortest, most resilient supply chains possible, while valorising waste and lowering costs.

“Cost reduction comes in both the R&D and production stages,” Leonard told Green Queen. “At the R&D stage, our platform reduces cost and timeline for ingredient discovery. Instead of relying on traditional methods, we utilise our AI to screen millions of molecules in sidestreams, their quantity, and predict their functions.

“The AI outputs a handful of ‘top candidates’ that we can then validate experimentally. Using AI, we reduce the amount of manpower and the number of ingredient candidates to analyse before identifying those suitable for our purposes. Ingredient discovery usually takes many years, but now it only takes months using the Portal and only a fraction of the cost.

“At the production stage, our parallel downstream processing hardware system tests multiple processes simultaneously to identify the most optimal one to maximise yield and quality. The hardware system of our platform circumvents the expensive and time-consuming traditional approach to developing a manufacturing process through sequential improvements.

AI can turn ‘cost centres’ into ‘resilient supply’ of healthier ingredients

According to 2nd Nature, the food system suffers from an “economic failure”, with millions of tonnes of nutrient-rich sidestreams treated as costly liabilities. It said the insights generated through the Kanematsu partnership could inform future discussions on innovation and sustainability opportunities.

“At Kanematsu, our corporate principle is to contribute to a sustainable society through innovation,” said Kazuhiro Matsuura, general manager of the Japanese corporation’s growth strategy office.

“Through this collaboration, we look forward to learning from 2nd Nature’s AI-driven approach and gaining new perspectives on how agricultural sidestreams may contribute to more sustainable resource utilisation,” he said.

The first step of the partnership is focused on Discovery, with 2nd Nature mapping the molecular potential of Kanematsu’s sidestreams to identify valuable compounds like proteins, functional fibres, and bioactives, based on their predicted functional properties.

“While our work with Kanematsu begins with mapping molecular potential, our platform is built to eventually help partners turn their biggest cost centres into resilient supplies of safer and healthier ingredients, as well as new revenue streams,” said Leonard.

The startup has already launched two products for food industry use: a zero-calorie sweetener for beverages, dairy and alternatives, baked goods, and snacks; and a non-sodium umami enhancer for plant-based meat, savoury snacks, ready meals, soups, and sauces.

“We believe that partnering with agri-food processors and CPG companies is the most efficient way for the industry to bring new ingredients to the market, whereby 2nd Nature can leverage its strength in ingredient discovery using agrifood solutions, and our partners can harness their expertise in distribution and market placement,” Leonard told Green Queen.

The company intends to file a Generally Recognized as Safe (GRAS) application with the US Food and Drug Administration. “We have an initial investment in our company, and currently fundraising for a seed/Series A round,” the CEO revealed. “We intend to use the proceeds of the round mainly for commercial expansion and making our AI tool even smarter.”

In another instance of an industry giant tapping into AI for ingredient discovery, Ingredion has teamed up with US startup Shiru to identify novel functional proteins for the food, beverage, supplement and specialised nutrition sectors, with an initial eye on the GLP-1 market.