SkinDeepRESEARCHSteve Seguin

Use case · Idea to test

Material research

Choose candidates under several requirements.

  1. Candidate materials
  2. Score the trade-offs
  3. Test selected candidates
Illustration of the proposed workflow.

A researcher might seek a material that balances strength, weight, conductivity or environmental cost. A small model could help choose which candidate deserves the next expensive evaluation.

Where the small model fits

A domain model represents the candidates; a trained scorer estimates the chosen properties or learns an expert's trade-offs. Known requirements remain separate constraints.

What would need to work

Count useful candidates found per evaluation budget, using independent measurements. The lab's toy constraint checks do not validate physical material properties.

Related experiments and evidence