Think of a number…

Does everyone understand what the AHI really means?

byTony McGrail

Think of a number…AHI McGrail
Image for illustration purposes

An asset health index (AHI) is a means to summarise a lot of disparate data into a single value to address a specific question: Does the asset need maintenance, for example, and if so, when? This short discussion will not go into the details of generating an AHI but will focus on what happens next (1).

So, the big question is: does everyone understand what the AHI really means?

AHIs are commonly used to rank assets for prioritisation of maintenance, replacement, or other intervention. There is a lot of work that goes on to reduce all relevant and available data to an AHI which covers the 4 key aspects of an AHI (2):

  • Calibration, so that the timescales for action of similar indices are common
  • Monotonicity, so that worse indices are always associated with a more urgent timescale
  • Auditable, so we can see which data was important in a specific index being generated
  • Justifiable, so that if we have to spend money, we have a good reason as to why ̶  we can see the failure modes identified and what they imply

Basically, the AHI summarises asset, condition, and operational data so that it may be used in financial and operational planning, as shown in Fig. 1.

The answer to the question “What does 5.7 mean?” is not a simple one: What would be a ‘good’ AHI value, or a ‘bad’ one?

What are the actions associated with that AHI? What are the timescales? What was the root cause data which yielded this score? How accurate is the analysis? And with what precision?

Whoever generated the AHI should have some understanding of all these questions – the AHI is a result of a technical evaluation of varied and imprecise data. What about the people using the AHI in analyses for planning? They usually understand numbers, spreadsheets, and equations – but rarely do they understand the actual assets themselves. The consequence, described as the Dunning-Kruger effect (3), is an overestimation of ability in a particular field of study based on limited knowledge.

The result is that the technical information used to derive an AHI is ‘lost’ as the ‘end user’ feels they know all they need to know as they understand the number, if not the asset. After all, anyone can see that 5.7 is bigger than 4 and therefore must be more likely to fail or need attention. The Dunning-Kruger effect has been found in many industries and applications. What we need to do is be prepared for it and to respond to it.



Log in or subscribe to read the whole article.

To read the article, subscribe and choose the option which suits you best. We offer both free and paid options, and the registration takes only a minute.
Subscribe to Transformers Magazine