Discovery of symbiotic opportunities in industry with AI
Most AI projects optimize for predefined answers. But how do you design an AI pipeline for problems without known solutions? In this webinar, you will discover how we used the concept of "undiscovered public knowledge" to uncover hidden opportunities for industrial symbiosis —the systematic conversion of the waste stream from one industrial process into the raw material for another.
Based on an analysis of more than 150 producers at two Belgian industrial sites, we have built a hybrid pipeline in which three different layers challenge and validate each other:
Public data: Deterministic Python rules that structure publicly available business information to define the playing field — which material flows exist, and which combinations can be ruled out in advance.
Domain expertise : Practical knowledge of material and upcycling specialists who assess every remaining combination against technical feasibility, regulations, and economic reality.
LLM reasoning: Large language models that mobilize knowledge from dozens of domains simultaneously to screen thousands of material combinations — not because the questions themselves are difficult, but because asking them manually via experts or literature becomes prohibitively slow at this scale.
What will you learn during this webinar?
- Alternative AI Design: Why open problem spaces require a fundamentally different AI approach than traditional operational automation.
- The 3-layer framework: How public data, domain expertise, and LLMs collaborate and address each other's blind spots.
- Results from practice: Concrete data and lessons learned from two active Belgian industrial sites.
- The reality of symbiosis: Why "plug-and-play" reuse is a myth and which systemic changes are truly needed to make material exchange financially and operationally viable.
Who is this interesting for
- Managers and advisors of industrial parks and business centers.
- Consultants in the circular economy and experts in material flow analysis (MFA).
- Professionals in waste collection, sorting, or secondary raw materials.
- AI specialists and innovation managers who want to use LLMs for complex, multivariate issues.
Practical information
- Date : Tuesday, October 15, 2026
- Program : * 10:00–11:00 — Methodology and results of the case studies
- 11 :00–12:00 — Open Q&A and discussion
- Language : Dutch (the slides are in English)
- Location : Online (the link will be shared after registration)
Participation is free, but registration is required to reserve your digital seat.
This research was conducted with the support of the Province of East Flanders, in collaboration with Material Mastery, BEPASIN, and the Aalst Business Association.