From Object-Centric Event Data to Causal Inventory Insights: A Structural Equation Model of Understock and Overstock

Abstract

Inventory behavior emerges from interacting items, locations, and orders, making case-centric analysis and purely correlational models ill-suited for intervention questions. We present an object-centric process mining pipeline that turns object-centric execution data into causal inventory insights. From an Object-Centric Event Log (OCEL), we reconstruct item-location inventory trajectories and derive interpretable indicators for demand, replenishment lead-time performance, ordering batchiness, and buffering posture. We then link these indicators to understock and overstock via a Structural Equation Model (SEM) with latent constructs, using time-in-state exposure as outcomes. The fitted SEM enables direct/indirect effect decomposition and lightweight counterfactual what-if analyses by propagating shocks through the model. We demonstrate the end-to-end approach on simulated OCEL-like data and on extracted OCEL measurables, providing transparent reasoning about the understock-overstock trade-off.

Authors 3

1. Alessandro Berti corresponding Aachen Process and Data Science (PADS)

   RWTH Aachen University

   Affiliation as printed

   Process and Data Science (PADS), RWTH Aachen University

2. Dina Kretzschmann Aachen Process and Data Science (PADS)

   RWTH Aachen University

   Affiliation as printed

   Process and Data Science (PADS), RWTH Aachen University

3. Wil M P van der Aalst Aachen Process and Data Science (PADS)

   RWTH Aachen University

   Affiliation as printed

   Celonis

   Process and Data Science (PADS), RWTH Aachen University