Data Driven Supply Chain Network Design
summary
Pipeline that ingests SAP/WMS and external rental cost data into an SSAS data model with geocoding via API, then runs weighted k‑means (scikit‑learn) to find centers of gravity and builds a linear programming optimization implemented as linear programming in Python to choose number, location and size of warehouses. Models include constraints (warehouse count, 24h fulfillment targets) and scenario sweeps; outputs are cost/service tradeoffs and ranked, human‑reviewable recommendations.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:13About Bosch and supply chain
- 2:28Supply chain network design solution
- 3:53Hierarchy of supply chain layers
- 5:30Typical project workflow
- 8:05Data sources and integration
- 11:30Center of gravity analysis
- 14:37Input definition and model development
- 16:52Linear programming optimization
- 18:39Scenario development and recommendations
- 20:06Evaluating recommendations
- 22:11Qualitative factors in decisions
- 22:58Human expertise in optimization
- 23:42Q&A
related talks
01h 18msupplyseer: Computational Supply Chain with Python
40mDomain-driven Data Science
28mContainer Tracking for Bosch Tools Supply Chain Operations
43mData warehousing reinvented for today's needs
23mData Team as an Optimisation Problem
35mData Warehousing with Python
32mScalable Python and SQL Data Engineering without Migraines
49mThe Data Supply Chain: Maximizing Value Throughout the Cycle