Requirement parser
Turns project briefs into application, material, performance, curing, and risk fields.
GeoAdditive AI Designer
A scientific design console for translating natural-language geotechnical requirements into structured material screening, evidence-backed candidate systems, risk notes, and experiment matrices.
Marine soft clay stabilisation with sulfate exposure, high early strength, and low-carbon boundary.
- Target: 7-day UCS and low permeability
- Materials: calcined clay, slag, fly ash, PCE
- Decision status: screening only, validation required
Recommended system
Strength potential
Carbon reduction
Cost control
Constructability
| ID | Variable | Levels | Tests |
|---|
Designed for auditable material decisions
The console is not a black-box claim generator. It keeps assumptions, evidence coverage, risk triggers, and validation actions visible so a research or engineering team can decide what to test next.
- Structured requirement parsing for soil, binder, exposure, performance, cost, and carbon context.
- Candidate ranking with explicit trade-offs rather than one unqualified recommendation.
- Experiment matrices that convert screening results into laboratory validation work.
Product modules
Standard functions support screening and reporting. Advanced-material workflows remain gated for licensed pilot or customer access.
Recommendation engine
Scores candidate materials and formulations against constraints, carbon, cost, and risk boundaries.
Evidence trace
Links recommendations to knowledge-graph records, literature, internal tests, and review status.
Report drafting
Produces technical summaries, validation plans, and decision records for research or pilot projects.
Demo video and pilot workflow
The demo poster is retained as a lightweight preview for deployments where the full video is published alongside release evidence. Pilot access should be agreed with clear data, privacy, and validation boundaries.
