Container Tracking for Bosch Tools Supply Chain Operations
summary
Model combines daily AIS snapshots (since Nov 2021) with static route features to predict vessel ETA for 2,000 trips from Asia Pacific to Western Europe, using XGBoost + Hyperopt, SQL Server preprocessing and Power BI reports. System retrains daily, forecasts up to seven days, and cuts SAP lead time errors from tens of days to a mean absolute error distribution mostly within ±5 days and a bias of ≈1 day; it correctly flags multi‑day delays (one predicted ~15‑day slip).
AI-written summary · about
chapters (8)
- 0:00Bosch supply chain challenges
- 2:10Supply chain management disruptions
- 6:28Container tracking initiative
- 11:53Modeling container routes
- 16:36Machine learning for arrival predictions
- 19:45Model performance backtesting
- 23:24Future system improvements
- 25:32Q&A
related talks
28mDriving Energy Transition With Data And Technology: A Perspective From VTTI
16mWatts Up? Leveraging AI & Data for Energy Management
45mOur way to a data-driven enterprise - A Journey by Bosch
25mData Driven Supply Chain Network Design
11mAWS On Air ft. BMW & BCG
18mUsing Transformer Models to Predict Demand at Scale in Production
27mOptimizing Omnichannel Order Fulfillment with AI and Advanced Analytics
25mAI-Driven Supply Chain Security: Using AI-BOM to Detect Hidden Vulnerabilities