University of Twente Student Theses

Login

Enhancing Forecasting and Inventory Control for Seasonal Spare Parts at NS

Sambeek, G.J. van (2024) Enhancing Forecasting and Inventory Control for Seasonal Spare Parts at NS.

[img] PDF
1MB
Abstract:This research at NS focuses on improving the inventory for seasonal spare parts, which experience higher demand during seasons. The Supply Chain Operations (SCO) department's process is outdated, relying on assumptions rather than structured models. The problems are the lack of specific forecast methods for seasonal parts and inventory policies that don't account for time-based changes, negatively impacting performance. SCO currently identifies 59 parts as seasonal, with unclear demand patterns and unknown lead times. Manual adjustments to forecasts often result in overestimations. Also, performance metrics like the fill rate exceed norms. The research reclassifies seasonal parts using multiple regression, finding 88 parts with seasonal patterns, of which only nine match the original list. Demand is forecasted using eight methods, with different methods performing best for various demand types: Croston for smooth, (S)ARIMA for intermittent and lumpy, and SES for erratic demand. The new model outperforms the current forecast process. Inventory is managed using the (s, (n)Q)-policy, with a fixed lot size preferred. Simulations show improved performance, reducing costs by 28.3%. Recommendations include biannual identification of seasonal parts, implementing new performance measures, and ongoing testing of forecast methods to optimize inventory management.
Item Type:Essay (Master)
Clients:
NS, Utrecht, Nederland
Faculty:BMS: Behavioural, Management and Social Sciences
Programme:Industrial Engineering and Management MSc (60029)
Link to this item:https://purl.utwente.nl/essays/101505
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page