An intervention-aware analytics workflow for estimating repair impact from hydraulic sensor logs

Key Points

• The research aims to estimate the effects of service actions on repair needs using hydraulic sensor logs.
• Utilized front-door adjustment with energy-balance mediators to identify causal service-action effects.
• Validated mediator overlap across leakage grades to ensure positivity before making inferences.
• Developed an analytics workflow to support maintenance decision-making based on hydraulic data.
• Established causal estimates indicating significant repair impact from specific service actions.
• Validated the energy-balance mediators with strong overlap across leakage grades, enhancing the credibility of findings.