Reliability Radio EP 329: IBM Predictive Maintenance misinterpreted -Tom Woginrich

ReliabilityRadio

Join Jonathan Guiney and Brendon Russ on Reliability Radio for a controversial debate with Tom Woginrich from IBM on "predictive maintenance vs. condition-based maintenance." Tom challenges the traditional view, defining the "predictive maintenance" of old as limited "univariate analytics" (single variables like vibration), often leading to frustration and missed failures.

He then unveils the true power of multivariate analytics: high-frequency data from dozens of variables (amps, temperature, load, even time of day) analyzed by machine learning to predict problems long before they become potential failures. Tom shares a fascinating anecdote of how a simple window shade caused a multi-million dollar recall for an automotive manufacturer, a problem only uncovered through disparate data analysis. Learn how IBM Maximo's capabilities, including multivariable health scores, are building the foundation for this next-level prediction. The discussion culminates with a glimpse into the future, where AI-powered systems automatically suggest actions without human prompting. This episode is a must-listen for anyone ready to move beyond guesswork and unlock the true predictive power of their asset data.