Recap: Co-creating the JASON AI tool at our first regional workshops

On Tuesday 28/04 and Wednesday 29/04 the ARGONAUT project hosted its first series of online workshops titled: “Measuring what matters: AI & Transparency in Bio-Based Products.” Over two interactive sessions, we brought together SME producers, researchers, policymakers, and industry professionals from across Europe to explore a critical question: How can technology make sustainability transparent, verifiable, and accessible?

The highlight of both events was the first-ever public demonstration of JASON, our AI-powered environmental impact assessment tool.

We would like to thank everyone that attended and provided valuable feedback for the next steps of JASON’s development. If you weren’t able to join us live, here is a recap of what happened inside our digital labs and the key insights we took away from our stakeholders.

Two Regions, One Shared Challenge

To ensure our tools reflect the real-world complexities of the bio-economy, we split the workshops to focus on two distinct European contexts.

  • The Italian Ecosystem (April 28): This session focused heavily on Bio-based Industrial Applications. Representatives from our Italian use cases shared their experiences in transforming agricultural and agri-food by-products into high-value biomaterials, green chemicals, and renewable energy. 
  • The Lithuanian Ecosystem (April 29): Our second session shifted the focus to Agri-food Based Bio-Products. Here, the conversation centered around regenerative farming, short supply chains, and upcycled food and fiber products.

Following the use case presentations, our technical partner from VCG took the virtual stage for a live demonstration of the JASON tool.

ARGONAUT workshop2
Screenshot from JASON's demonstration

JASON is designed to do the heavy lifting of environmental assessment, using Large Language Models (LLMs) to scan a large database of scientific papers, estimate missing data points (imputation), and generate accurate environmental footprints for bio-based products. During the demonstration we asked the audience to “co-create” the tool’s future through interactive polling and open discussions. Here are four key takeaways from that feedback:

  1. The Demand is High, and the Interface is on the Right Track: There was an overwhelming consensus among participants that conducting environmental impact assessments for bio-based products is an absolute necessity. Fortunately, the initial reaction to JASON was highly positive. A strong majority indicated they would be willing to adopt and engage with a solution like JASON.
  2. SMEs Need More “Hand-Holding” During Data Entry: While users understood the information being asked of them, there were mixed feelings about whether it is realistic for a typical bio-based producer to complete all the fields. Participants pointed out that open-text boxes can be confusing without strict parameters. The most requested improvements for the final version include adding pop-up tooltips, examples of correct answers, clearly defined units of measure, and distinguishing between mandatory and optional fields.
  3. The “LCA Expertise” Barrier Remains a Concern: A recurring theme in the feedback was the concern over data availability. Participants noted that some of the required fields are highly technical and might still require specific Life Cycle Assessment (LCA) expertise to answer correctly. There is a worry that smaller SMEs might not have access to the detailed composition or process data that the tool asks for, highlighting the need for JASON to potentially offer default values, drop-down menus, or partial auto-fill options for non-experts.
  4. “Trust, but Verify” the AI: When asked if they would trust an environmental assessment made with the assistance of an AI tool, the response was generally positive but still cautious. Participants agreed they could use this information for business and policy decision-making, provided the tool is completely transparent. Stakeholders want to see exactly where the data is coming from, what benchmark data is being used, and even flowcharts of how the calculations are made to ensure the results are 100% credible.

What’s Next? The Road to October

The insights gathered from these workshops are already being fed back into our development pipeline. Over the next few months, our technical teams will refine JASON’s user interface and algorithmic accuracy based directly on your feedback.

In October, we will be hosting our second round of workshops. Not only will we unveil the updated, fully functional version of JASON, but we will also showcase a first version of the Digital Bio-product Passports (DBPs). We will show exactly what happens when a consumer scans a product and sees an AI-calculated, blockchain-verified sustainability profile.

Thanks again to everyone who participated, challenged our assumptions, and contributed to the future of the circular bio-economy.

Don’t miss the next update! Make sure to subscribe to the ARGONAUT newsletter and follow us on social media to stay updated!

Share this article:

Subscribe to our newsletter​

Follow us: