Tyler Fedun

In early 2025, Energy Transfer Partners rolled out a machine learning enabled predictive maintenance program across their extensive pipeline network. One year into their five year journey, Mike Martin will share both the promise and the practical challenges of scaling predictive analytics across their fleet of complex rotating equipment systems.

• How a centralized data infrastructure enables successful machine learning and why it is the critical first step
• Lessons learned from implementing predictive models, including early successes using historical failure scenarios
• Leveraging operational data to proactively detect failures, identify anomalies, and optimize maintenance strategies
• Strategies for scaling from pilot programs to enterprise-wide deployment across hundreds of assets
• The roadmap toward true predictive maintenance, reducing reliance on calendar-based and runtime-based servicing