Thoughts, tools, and field stories from the intersection of reliability engineering, predictive maintenance, and AI technology.
We fed an Outotec P0668 31m thickener SAP hierarchy into an AI agent built on Claude, applied John Moubray's RCM2 decision logic as a structured prompt chain, and got 64 failure modes with full three-level effects, risk scoring, and a PM task register — in a single session. Here is what it means for reliability engineering practice.
A deep dive into the three pillars of predictive maintenance and how they reveal equipment health before failure hits.
How machine learning models are transforming maintenance strategies — and the critical pitfalls when implementing them in real industrial environments.
Why intuition, experience, and human judgment remain essential in predictive maintenance — and how to design systems that enhance rather than replace them.
Moubray was clear: consequence before task. Skip that order and you are not doing RCM2 — you are doing traditional maintenance with better paperwork.
Stories of success, failure, and everything in between from 15 years of real industrial environments — and the insights that stayed with me.
The functional location hierarchy is the backbone of your SAP PM system. Get it wrong at the start and you spend years working around it.
My PhD research is answering this question empirically. Here is the state of the debate and where the evidence is pointing.
Suggest a topic, share a field story, or ask a reliability question worth writing about.