S11-U. Cold-Chain · Logistics SLA + Store Ordering (UPI-Specific)
Published 2026-05-14Updated 2026-06-302 min read
Uni-President differentiator ⭐ — the in-house logistics arm (President Transnet) runs the cold chain that directly determines quality for core SKUs (dairy, ice cream, bento). Combines outdoor temperature + store-order optimization.
1. URL · Persona
/cold-chain· P5 (Manufacturing · Logistics · Stores)
2. User Story
P5 — At southern heatwaves of 35℃+, derive the dairy cold-chain SLA breach rate + store-order correction recommendations.
3. Data Mix
| Data | Source |
|---|---|
| ColdChainSLA logs | TimeStream (~50K) |
| OwnSKU (cold-chain scope) | Neptune |
| Store (7ELE / Carrefour stores) | Neptune |
| Outdoor temperature | WeatherSignal |
| Store inventory | Inventory |
4. Processing Pipeline
1. Aggregate daily cold-chain SLA per (origin BU, destination store)
2. Join outdoor temperature (Central Weather Administration, 中央氣象署)
3. Count breaches (actual exceeds target_temp + 2.0℃)
4. Compute store-inventory remainder + breach impact
5. Simulate next-day ordering correction (Sonnet)
5. Output UI
- Left: SLA breach-rate time series per region (north / central / south)
- Center: Outdoor temperature vs breach-rate scatter plot
- Right: Per-store SLA-impact heatmap
- Bottom: Next-day ordering-correction simulation (inventory · forecast orders · cold-chain recovery)
6. Guardrails
- Block external exposure of store unit pricing and inventory
- Cite external weather sources
7. Demo Scenarios
- Southern heatwave 35℃+ → dairy cold-chain breach rate 28% (vs. 8% baseline)
- Next-day ordering correction → -20% orders to breaching stores, +10% to stable stores
- Weekly ordering simulation → recovery cost per region · SKU