Causal inference has transformed science in recent decades by addressing questions of causality (“why?”). This research project aims to apply causal machine learning to prescriptive machine maintenance, which currently relies predominantly on associational learning. By identifying and quantifying causal relationships between (preventive) maintenance decisions and machine failures, we investigate how the type and timing of maintenance actions affect future breakdown. To do so, we account for equipment characteristics, failure mechanisms, maintenance history, and operational sensor data. The project’s ultimate goal is to develop decision-support models that enable more effective maintenance interventions.