Business Objective
Our client is the fleet planning department of a leading operator in the North American railroad industry with a fleet of 200,000+ railcars.
The client wanted to develop an automated railcar fleet planning system to enable capacity planning and maintenance planning.
Challenges
- Large number of internal and external Data sources
- Complexity of business operations
- Manual and ad hoc data collection/preparation
- Integration of data automation and analytical modeling layers
Solution Methodology
- Developed models to forecast Import/Export commodity volumes for the United States and Canada based on import/export data, industry data, and macroeconomic data. Automated data extraction, preparation, and modeling
- Developed a suite of models to forecast Loadings for each O-D (Origin-Destination), Lane by Fleet Type using Import/Export forecasts as inputs.
- Used the loading forecast model as input to develop models for forecasting Cycle Time for each O-D Lane by Fleet Type, monthly mileage for each fleet type, and fleet demand/usage
- Developed models to forecast demand for each fleet type in the North American market. Developed business logic for capacity planning and placing new fleet orders
Business Impact
- Developed a fully automated system to generate fleet forecasts, identify future capacity requirements, and place orders for new railcars.
- The system:
- Eliminates manual steps and automates data extraction and preparation from numerous data sources
- Integrates data automation layer with an embedded analytical modeling layer
- Has “What-if” analysis capability to analyze different economic scenarios e.g., growth, decline, neutral, etc
