Improving safety and operational efficiency in the rail transport industry relies on a precise understanding of the root causes behind system failures. In this research, we propose RaiLog RCA, a comprehensive root cause analysis approach leveraging log data from railway operating systems. This approach is designed to detect the Point of Incipient Failure through the analysis of real-world time-series data. By constructing a structural causal model and applying probabilistic counterfactual analysis, RaiLog RCA provides actionable insights that enhance the root cause discovery from identification of the time of appearance of anomalies and the associated causal graph. The paper presents the experimental results that demonstrate the accuracy performance of the proposed approach.