JapanRetail and e-commerce2023 – ongoingSuper Hosokawa
Shared demand forecasting trial across the supply chain
Japanese supermarket trial reports 0.20% tofu and fried-product waste with combined ordering, versus 0.52% overall
Problem
Businesses create over half of Japan's food waste, and each tier of the supply chain forecasts demand on its own data. The result is overstocking up the chain.
Approach
Three-store operator Super Hosokawa and six other companies shared ID-POS data, which flowed through Azure Blob Storage and Databricks into a 60-pattern forecast model. Forecasts ran two days ahead and store staff combined them with their own orders. The trial began in January 2024.
Results
Takeaway
Combining forecasts with staff ordering decisions produced a 0.20% waste rate for tofu and fried products in the trial, compared with 0.52% overall. Some projected results remain unconfirmed.
Categories
- Size
- Small business
- Industry
- Retail and e-commerce
- Business function
- Logistics and supply chainProduction and quality
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Vendor solution
- Evidence
- Vendor's case study
Sources
Source summaries
Microsoft reports a supply-chain demand forecasting trial at a three-store Japanese supermarket. It lists forecast error and food waste results.
Full details are in the original sources.
