Federated Digital Twin Sandbox & What-If Simulation

The Digital Twin Causal Sandbox is SweetNight DXP’s “flight simulator”. Grounded in Structural Causal Models (SCM) and Bayesian networks, it maps real-world operations across the 1+6 towers into a virtual simulation environment. Executives can simulate black swan scenarios, test extreme cost pressures, and formulate hedging strategies without risking live capital.

DXP Digital Twin Causal Simulation & Scenario Branching

5 Macro Sensitivity Sliders

The sandbox models five exogenous and endogenous drivers affecting cross-border home furnishing operations:

Sensitivity Slider	Symbol	Range	Live Baseline	Operational Entity & Impact
01 · Commodity Raw Material Inflation	x_1	`-15.0% ~ +25.0%`	+8.2%	Drives polyol foam costs, spring steel wire pricing, and standard BOM costs.
02 · Commercial Price Tuning	x_2	`-10.0% ~ +10.0%`	+2.2%	Evaluates LightGBM discrete elasticity, predicting demand volume and unit margins.
03 · Ocean Freight per FEU	x_3	`1,200 ~ 4,500`	2,850	Determines 40HQ landed costs, port demurrage, and forward warehouse storage rates.
04 · USD / CNY Exchange Rate	x_4	`6.80 ~ 7.40`	7.22	Influences FX conversion spreads between US dollar revenue and CNY factory procurement.
05 · Paid Ad Blended ROAS	x_5	`2.50x ~ 6.00x`	4.35x	Affects marginal acquisition costs (CAC) across Shopify DTC and Amazon PPC campaigns.

Causal Mathematical Equations


ΔRevenue = Base Revenue × (1 + x₂) × (1 + ε · x₂)
ΔCOGS = Base COGS × (1 + 0.65 · x₁) × (x₄ / Base FX)
ΔFreight = FEU Volume × x₃ × (1 - Rollpack Yield Gain)
ΔAd Spend = ΔRevenue / x₅
Projected Net Margin = (ΔRevenue - ΔCOGS - ΔFreight - ΔAd Spend - Opex) / ΔRevenue

Where:

ε = -1.17 denotes the demand price elasticity predicted by LightGBM.
0.65 represents the direct material weight of polyol and steel in total manufacturing costs.
Rollpack Yield Gain = 0.276 captures the freight dilution achieved by high-tonnage hydraulic packaging.

Bayesian Goal-Seeking Inverse Solver

Traditional BI systems only support forward calculation (“given causes, compute effects”). DXP introduces an Inverse Solver allowing leadership to declare a target outcome and solve for required operational inputs:

Inverse Optimization Scenario:

Target: Sustain a contribution margin of 23.50% despite raw materials at +8.2% and freight at 3,200/FEU.

Clicking “Inverse Solver” activates optimization routines that converge on optimal parameters:

Commercial Price Tuning (x_2): Raise by +2.8% (volume softens by 3.2%, but revenue expansion covers inflation).
Paid Ad ROAS (x_5): Target 4.60x on Shopify DTC (retire low-efficiency keywords, prioritize high-intent long-tail).
Packaging Density: Increase Vietnam factory roll-pack compression by +3.5%, fitting 595 units per 40HQ.
Result: Settles margin at 23.52%, expanding net cash reserves by 340,000.

3 Pre-Configured Stress-Testing Playbooks

1. Playbook A: Black Swan Supply Shock

Inputs: Raw materials +20.0%, Ocean freight surges to 4,200/FEU.
Alert: Without hedging, net margin breaches the 15.0% insolvency red line.
Hedging: Activate alternative foam supplier certifications and stage tiered price increases.

2. Playbook B: Aggressive Price War

Inputs: Primary competitor drops price by 15.0%, eroding Buy Box win rate to 62.0%.
Diagnosis: Price matching causes an 850,000 monthly loss; recommend defending flagship models while introducing secondary value bundles.

3. Playbook C: Currency Appreciation

Inputs: Exchange rate plunges from 7.22 to 6.85.
Hedging: Execute forward FX swap hedges and transition supply contracts to RMB billing.