Interpreting data – context is key

We’ve told you before; data is the new gold. In the new digital world we are constantly generating- or protecting data. Today's technology enables you to measure everything. From your heart rate to your sleep rhythm and from your search engine behavior to your Netflix preferences. To measure is to know! But, is it really necessary to measure everything? And is just measuring enough to know?

Why would you want to measure something?

It all starts with this question. Generating data for the sake of generating data is completely pointless. Raw data alone does not tell you much. It is therefore important to consider in advance what it is you want to find out from the data. Think carefully about what you want the data to tell you and - perhaps even more important - in what context you are going to place this data.

The value of context

Let's give you an example: suppose your heart rate monitor indicates eighty-five beats per minute. At that moment you only know that your heart beats eighty-five times within the time span of one minute. This fact alone says nothing about your health, or your fitness. You might even be shocked by it, thinking it might be way too high.


It only gets interesting when you add context to this data: are you at rest, or straining yourself? How old are you? What is your gender? This will get you a bit further, but actually you still can't draw any conclusions.
One more ingredient is needed; peer groups, comparisons. What is the heart rate of people of the same sex, the same age, who make the same effort as you? Compare your data with that and only then will you know something about your fitness; then your data becomes valuable.

Data + context + peer group = value

 

How can you best interpret data?

In addition to context and comparison material, more is needed to interpret data correctly. First of all, it is important to check how reliable the data is; ask yourself how the data was measured, how accurate it is and whether there are chances of deviations. In the example of the heart rate monitor; is it attached in the right place on your body and have you worn it long enough to measure a reliable heart rate?

Furthermore, it is important that you select for relevance. You don't have to measure everything to draw a conclusion. Measurements sometimes give much more information than is relevant and necessary. By selecting relevant data you distinguish between main and secondary issues. This prevents noise on the line. If we want to know how quickly the doors of an elevator wear out, it is especially important to measure the number of door movements. How often the elevator moves up and down the elevator shaft is irrelevant here.

'I install an elevator sensor and I'm ready for predictive maintenance'

So while many see an elevator sensor as an interesting gadget, it doesn't quite work that way. There is more to working with predictive maintenance. It works exactly the same as the heart rate monitor.
A sensor is needed to generate data, but it is up to you what you will do with that data. By determining in advance what you want to learn from the data generated, placing it in context and comparing it with data from other elevators. This way, patterns can be identified and predictions made regarding malfunctions and maintenance. Our software helps you put the data measured by our sensor in context and compare it with other data.

Want to know more about elevator data or predictive maintenance? Please feel free to contact us!

Dit bericht delen: