Karthik Sekar, head of data sciences at Climax Foods explained to FoodNavigator-USA the Benchling partnership will accelerate the company’s process to develop “tasty, nutritious and affordable plant-based cheeses that haven’t been seen before.”
“Our team uses data science to find the most expedient path toward better plant-based products. Our methods are faster than traditional trial-and-error methods, which might otherwise take decades to reach the same outcome,” Sekar emphasized.
In April this year, Climax Foods announced its partnership with BabyBel maker, The Bel Group, as reported previously by FoodNavigator-USA. The goal of the collaboration is to produce more affordable plant-based cheeses formulated with a near identical taste, texture and melting characteristics as its animal-based counterpart. Further Bel Group intends to "achieve...a group goal of a 50% plant-based product line in ten years," Sekar noted of the partnership and its impact on plant-based cheese development.
"Machine learning models could potentially predict novel properties of such ingredients"
Sekar continues to explain that AI’s capabilities, compared to a traditional and often resource-intensive R&D process, can contribute to a more efficient discovery and use of novel plant ingredients that are selected to enhance flavor, functionality and sustainability of a plant-based product.
“There are an estimated 300,000 edible varieties of plants in the world. A given grocery store may make a mere 100 available. Examining this untapped space of plant ingredients may proffer new, enhanced flavor, functionality, and sustainability. The sheer quantity of edible plants cannot be investigated using traditional lab science–there is not enough time, personnel, nor funding to deploy. Instead, machine learning models could potentially predict novel properties of such ingredients without having to go into a lab.”
While ethical implications on AI-use continue to raise questions around intellectual property, legal procedures, job security and false information, among others, Sekar emphasizes the technical capabilities of machine learning towards plant-based food development, allowing scientists and stakeholders to work within a streamlined process.
“These foods must still be physically created by individuals and consumer feedback will continue to be collected. Novel ingredients will undergo all appropriate approval processes. All and all, the specificity of our machine learning and the continual involvement of scientists, consumers, and regulatory bodies ensure that any outsized risks are minimized.”