SciTransfer
COCO-AI · Project

AI-Driven Lab-Grown Plant Ingredients for Sustainable High-Value Food and Cosmetic Production

foodPilotedTRL 7

Imagine growing the best parts of a plant in a giant steel tank instead of a field. It's like brewing beer, but instead of alcohol, you're making rare cocoa ingredients without needing hectares of land. An AI acts as a master chef, constantly tweaking the recipe to make the process faster and cheaper.

By the numbers
10,000L
bioreactor scale
6
novel secondary metabolite formulations
2
prototype chocolate bars
The business problem

What needed solving

Companies rely on unstable and unsustainable imports of plant-based natural products. Current lab-grown alternatives are too expensive and slow to develop for mass market use.

The solution

What was built

An AI-optimised plant cell culture platform, 6 metabolite formulations, 2 prototype chocolate bars, and open-source AI tools.

Audience

Who needs this

Chocolate and confectionery manufacturersBio-based cosmetic producersPlant-based pharmaceutical companiesAgricultural biotech SMEs
Business applications

Who can put this to work

Confectionery
enterprise
Target: Chocolate manufacturer

If you are a chocolate manufacturer dealing with fragile cocoa imports and unstable supply chains — this project developed a plant cell culture platform that produces cocoa secondary metabolites at a 10,000L scale. This allows for consistent ingredient sourcing without relying on terrestrial farming.

Cosmetics
SME
Target: Bio-based skincare brand

If you are a skincare brand dealing with the high cost of rare plant extracts — this project developed AI-optimised culture conditions that reduce development costs. You can produce high-value secondary metabolites in a controlled environment.

Biotechnology
mid-size
Target: Specialty chemical producer

If you are a chemical producer dealing with slow R&D cycles for bio-based products — this project developed open-source AI tools and validated SOPs. This accelerates the transition from lab to industrial production for various plant species.

Frequently asked

Quick answers

What is the industrial scale of this production?

The project aims to scale production from the laboratory level up to 10,000L bioreactors.

How does this affect production costs?

The integration of an AI agent across the bioproduction workflow is designed to predict optimised conditions, thereby reducing development costs.

What is the IP and licensing situation?

The project will generate open-source AI tools and publicly accessible datasets to lower entry barriers for SMEs.

Are there regulatory safeguards in place?

The project includes regulatory and consumer studies to guide a safe and sustainable market introduction.

When will the results be available?

The project period runs from 2026-05-01 to 2030-04-30.

Consortium

Who built it

The consortium is highly commercially oriented with a 44% industry ratio, comprising 4 industrial partners and 5 SMEs. With 9 partners across 7 countries, the group combines the applied research power of Fraunhofer with international expertise from Canada, Israel, South Korea, Lithuania, the Netherlands, and the UK, ensuring a strong path toward market adoption.

How to reach the team

Contact Fraunhofer Gesellschaft for details on the AI-optimised platform

Next steps

Talk to the team behind this work.

Contact us to identify licensing opportunities for the open-source AI tools.

More in Food & Agriculture
See all Food & Agriculture projects