hardCase StudyDesignMidSeniorStaff
You have three years of vibration data and eleven recorded failures. What can you honestly build from that?
7 min readUpdated 2026-08-15Target archetype: Enterprise Captive, Product Startup
Answer
Short answer
Eleven events will not support a supervised remaining-life model, and evaluating one row-wise will make it look excellent. Build a per-machine deviation model with physics-derived features, and score it on how many of the eleven it caught with useful warning and how many alerts that cost per machine per month.
Was this answer helpful?
Likely follow-ups
- The work order says the bearing was replaced. How do you establish when it actually started to fail?
- Two of the eleven happened on the same machine. What does that do to your split?
- Your model catches nine of eleven with six weeks' warning. Is six weeks useful, and to whom?
- What would you build first if there were no recorded failures at all?
Related questions
- A churn model scores 0.94 AUC offline and barely beats the baseline in production. One of its strongest features is populated by a process that runs after the customer has already churned. Explain what happened and how you would fix it.mediumAlso on model-evaluation5 min
- Your predictive maintenance system is live and the maintenance team has stopped acting on its alerts. How do you get that back?hardAlso on predictive-maintenance6 min
- Most predictive maintenance projects never reach production. Why, and what would you do differently?hardAlso on predictive-maintenance6 min
- Design a notification system that sends 50 million messages a day across email, SMS and push.hardSame kind of round: design5 min