Optimization Parameters
Cost Weight (α)0.50
← GreenerCheaper →
Annual Demand (units)
Total units procured per year. Assumed to be split equally across all selected suppliers. Capacity constraints and minimum order quantities are not modelled in this prototype.
50,000
Demand Volatility (log-normal σ)
Standard deviation of ln(demand). Used in Monte Carlo simulation. Approximates the coefficient of variation (CV) for small values (<30%). E.g. σ=0.20 ≈ 20% CV.
0.20
Low riskHigh risk
Max Lead Time (days)30
Carbon Price ($/tCO₂e)$45
Max Suppliers4
Adjust parameters and click Run Optimization to generate the Pareto frontier and routing recommendations.
Supplier Network
Weighted-Sum Pareto Optimisation
For each α ∈ [0,1] (40 steps), suppliers are scored as:score = α·C̃ + (1−α)·Ẽ where C̃ and Ẽ are min-max normalised cost and emissions. Top-n suppliers by score are selected. Non-dominated solutions across all α values form the Pareto frontier.
Baseline Definition
The "baseline" is a naive equal-weight split across all eligible suppliers (those within the lead time constraint). It represents a simple unoptimised allocation — not a cost-optimal or emissions-optimal route. The number of eligible suppliers updates dynamically with the lead time slider.Monte Carlo Simulation
Demand D ~ LogNormal(μ, σ) where μ = ln(D̄) − σ²/2 and σ is the log-normal standard deviation set by the slider. 1,000 trials are drawn using the Box-Muller transform. VaR is the empirical 95th/99th percentile of the cost distribution.Carbon Price Sensitivity
Total cost = procurement cost + CO₂e × carbon price. The crossover point is where green routing total cost equals cost-optimised total cost as EU ETS price rises.| Supplier | Region | Unit Cost | CO₂e/unit | Lead | Reliability | Tier | Status |
|---|