BARCELONA — Nestle SA has accelerated its product development process by 60% since 2016, according to the company. The faster speed to market has been achieved through a restructuring of its research and development process. Now the company is investing in various forms of artificial intelligence (AI) and machine learning to further improve its R&D process and generate better results.

“We moved, actually, from an average project duration of 33 months to 12 months, and that's an average of different categories,” said Stefan Palzer, chief technology officer, during Nestle’s investor day on Nov. 29. “In food and beverage, sometimes projects take us only six to nine months, so we are faster now than many of the startups that are out there.”

The initial accelerated timeline was developed through simplification of Nestle’s overall R&D process. Project approval was reduced from six “gates” to three, the company established 14 R&D accelerators around the world where product development only takes six months from “idea to shop,” Palzer said, and product testing is done in real-world conditions.

The use of AI is spreading throughout the organization and is now used in a variety of ways, including concept development, formulation development, plant breeding, clinical data mining, raw material quality assurance, advanced process control and early problem detection.

“We established an artificial intelligence concept engine, which is transforming (social media) insights into concept proposals, which are then evaluated by our employees,” Palzer said.

Concepts that are approved are then prototyped and tested. As part of the prototyping process, another AI module that streamlines the formulation development process may be used.

Palzer emphasized AI and machine learning are now necessary product development tools to address the growing complexity of the product development process where products must taste good, be perceived as healthy, be sustainable and be affordable.

“To deal with this complexity we need artificial intelligence tools,” he said. “For instance, we developed a clinical data mining approach that allows us to do new discoveries based on existing clinical studies. So, we valorize much more (of) what we have already done in terms of clinical studies, and we use that to create new discoveries and new inventions.”

Beyond product development, Nestle also is using AI to improve manufacturing efficiencies. Palzer said some of the company’s KitKat manufacturing lines are self-regulating.

“We have loops in those lines, so they detect the product quality, they measure the attributes of the wafer, for instance, and then they regulate the process back,” he said. “So, it’s a self-controlling mechanism. But we have also developed, for instance, machine learning approaches for preventive maintenance. So, that allows us to reduce downtime of our lines.”