1 · Point-to-point graph
Every element is a node: origin point, transform point, install point, recovery point. Edges carry mass, distance and energy. The twin solves the lowest-mass path between the design intent and a certified supply node.
The AI Digital Twin top model, formula by formula.
Nothing here is a black box. The twin is a point-to-point graph solver wrapped in a material-substitution model; these are the exact quantities it returns for every API call.
2 · Baseline vs twin mass
A linear baseline mass Mb is computed from standard sections and virgin sourcing. The twin returns Mt, the mass of the optimised path after topology reduction, reuse substitution and right-sizing.
3 · Savings & multiplier
Material saving S = Mb − Mt (tonnes). The efficiency multiplier is X = Mb / Mt, which reaches up to 100x when a whole assembly is replaced by recovered stock plus a minimal connector set.
4 · Circular economy points
Points P = S × Cf × Rq × Lf. Cf is the stream carbon factor, Rq the recovery quality of the matched stock (0–1), Lf the loop factor (1.0 closed loop, 0.6 open loop, 0.3 downcycle).
Stream coefficients
| Stream | Carbon factor Cf | Recovery quality Rq | Basis |
|---|---|---|---|
| Critical minerals | 12.4 | 0.82 | Li, Co, Cu, rare earths |
| Circular plastics | 3.1 | 0.74 | Polymer-sorted flake |
| Circular textiles | 5.6 | 0.68 | Fibre-to-fibre yield |
Try the calculation
S = Mb − Mt · X = Mb / Mt · P = S × Cf × Rq × Lf
- Material saved
- 102.0 t
- Efficiency
- 6.7x
- Circularity points
- 1037
The same numbers are returned by /v1/twin/optimize and metered against the stream key that authenticated the call.