Evaluating Structural Drivers of Inventory Imbalance : Case Study in Rolling Stock Maintenance Company

Read full text Summarise with VIRAT
Authors: Topias Härkönen
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Persistent issues with inventory overstocking and shortage risks remain common in asset-intensive organisations that rely on MRP planning systems, particularly after system transitions. Prior research has mainly focused on addressing inventory imbalances through demand forecasting and service-level optimisation, often overlooking the role of governance planning parameters and master data integrity. This thesis treats inventory imbalance as a governance issue instead of a forecasting mistake. It uses historical ERP transaction data from a large organisation with significant assets. The study creates a framework to assess the risk of overstock and shortages at the item-week level. Overstocking and shortage risks are recognised as distinct and uneven outcomes. These outcomes are examined separately with clear logistic regression models that consider factors like demand patterns, supply uncertainty, batch limits, and data accuracy. The reliability of these models is tested through sensitivity analyses, multicollinearity checks, stability tests over time, and a non-linear Random Forest validation. Results show that ongoing inventory imbalance is more closely related to governed planning parameters, batch rigidity, and master data errors than to demand uncertainty on its own. It is harder to correctly classify overstock risk, and it often increases through specific planning interactions. In contrast, the shortage risk shows clearer structural patterns. Simply improving forecast accuracy will not fix the imbalance if planning parameters and data quality are not in sync. To tackle this, a composite governance risk indicator is suggested to transform item-level diagnostics into portfolio-level oversight. This research adds to existing literature by viewing inventory imbalance as a governance issue and showing how effective transparent, driver-based models can be as diagnostic tools within ERP environments. In practice, the study highlights the importance of governance after ERP transitions, focused master data audits, and tailored strategies to manage risks related to surplus and availability in spare parts inventories.

Fields of Study & Subjects

Supply Chain and Inventory Management ERP Systems Implementation and Impact Management and Optimization Techniques

Publication details

Journal / source
Doria (University of Helsinki)
Publication date
1 January 2026
Language
EN
Document type
Thesis