Predicting the Unpredictable:

How to Forecast Reliability with Limited Data and Unknown Configurations

At this year’s R.A.M.S. Europe Conference, our very own Elly van den Bliek challenged a long-standing belief in reliability engineering:
That meaningful forecasts require full knowledge of system architecture and a mountain of data.

In reality?
You can do much more with much less—if you know how to look at the right signals.

The problem:
Modern systems are complex, ever-changing, and often poorly documented. Failures in the field? Sporadic. Data? Limited.
Yet engineers and decision-makers are expected to predict reliability and make critical design and maintenance choices.
How?

That’s the question Elly tackled on stage.

The approach:
Elly demonstrated how combining Crow-AMSAA (for trend analysis) with Weibull models (for degradation insights) creates a surprisingly robust forecasting method—even when failure mechanisms are not yet fully understood.

No need for a full system map.
No need for decades of data.
Just a smart, structured way of working with what you do have.

Her approach included:

• Using trend lines to detect reliability growth (or decline)
• Applying Weibull analysis to anticipate emerging failure types
• Combining models to reinforce predictive power
• Supporting real-world decisions on lifetime extension, maintenance, and spare parts

Why it matters:
Elly’s message cuts to the heart of what we believe at Holland Innovative:
🚀 Reliability starts early.
Not when failures hit the field, but when systems are still being shaped.
Even if you don’t have the full picture—reliability thinking helps you avoid surprises, plan smart, and extend system life.

Want to learn how to do this yourself?
Our fall reliability courses teach you exactly these principles. Whether you’re just starting or ready to specialize, there’s a module for you:

• RF1 Life Data Analysis & Reliability Testing – 8 Sept
• RF2 System Reliability – 11 Sept
• AI for RCA (NEW)- 22 Sep
• AI for RCA (NEW) -3 Nov
• Workshop Root Cause Analysis – 11 Nov
• Workshop Reliability – 19 Nov
• RF9 Physics of Failure – 20 Nov
• AI for RCA (NEW)- 8 Dec
• RF6 Software Reliability – 1 Dec

🔗 Or dive into the full Reliability Foundation Program in our blog here.
📞 Want tailored advice? Let’s connect you with an expert.

Conclusion:
R.A.M.S. Europe 2025 reminded us again: true reliability engineering is about foresight, not hindsight. With the right tools—and the right mindset—you don’t need the full blueprint to build strong, data-informed decisions.

Let’s keep learning, sharing, and raising the bar for system reliability—together.