Monitor the right thing – Part 1 of 6 Ways to improve your I&M

What to do with data?

Data is incredibly useful in the construction and infrastructure world. We deal with complex, large-scale projects where control and decision-making are key. Data can help stakeholders make the right decisions at the right time. However, the key to using data for decision-making is that the data needs to be useful in the first place.

How often do we hear in the business world, don’t collect data unless you are going to do something with it? Or I don’t have the full dataset to understand this fully.

It might sound straightforward, but deciding on what to monitor, when to monitor, and why to monitor is often missed. As humans, we often jump straight into action—but as engineers, we strive to channel that impulse through curiosity, practicality, and a drive for solutions.

However, before even looking at the type of sensor, we must first answer one very simple question. This question was not one I came up with, that credit goes to the late John Dunnicliff. Having had the honour of attending his courses, one question was always asked first:

Seems simple right? This question is so often missed or misunderstood. It’s not what sensors do I need to install? It’s not, where do I need to install them?

Monitoring is about creating clarity, it’s about getting more understanding about something not fully understood, and therefore starting with specifying what the question we are trying to answer is the foundation.

Identify the parameters

Once we understand the questions that need answering, we can then look at what parameters we need to understand to answer that question. Often it is a collection of parameters that are used to build up a richer picture to be able to answer our geotechnical questions.

For example, when you visit a doctor with a fever, the question we are trying to answer is “What is causing that fever” or “Should I be worried about this fever”. We then carry out a set of tests, observations, or “monitoring” to create more clarity to answer the original question. Is the answer always immediate? No, that’s when further tests or expertise are required. Monitoring is no different.

Settlement monitoring

As an example in our industry, often the requirement to monitor is referred to as “settlement monitoring”. Whereas settlement is a symptom, which helps us answer a geotechnical question.

The question is “Has this ground consolidated enough to start construction”. If we zoom out on this question there could be a number of parameters that are useful to help answer this question:

1. Settlement – As before, this tells us exactly how much settlement has happened. However, it doesn’t tell us why, or why not. Also, settlement monitoring is not always total settlement, differential settlement is often important.
2. Pore water pressure – If the consolidation is too fast or too slow this could be affected by the pore water pressure. Knowing this information could also allow you to control the speed of the consolidation.
3. Deformation – As the ground is settling, does it also cause rotational movements, is it settling uniformly?

As you can see, by setting out the question you are trying to answer, there may be multiple parameters to help you understand the real uncertainty, not simply “settlement monitoring”

Specify the parameters

Now we understand the questions we are trying to answer and the parameters we need to measure, it’s time to add more detail to those parameters to ensure we get the correct data to answer our questions:

1. Can it be measured directly? – Not all parameters can be measured directly. We may need to measure a group of parameters or a proxy.
2. What is the expected rate of change? – You wouldn’t use a tape measure to measure the width of a hair. It’s important to know how much you expect the parameter to change to ensure it is measured effectively.
3. Is it cost-effective? – You might have the perfect parameter to monitor but if it’s too expensive to monitor for your specific application it may not make sense to measure. Is there another parameter or group of parameters that can be measured more cost-effectively whilst still answering your question
4. How often does it need monitoring? – This can affect the methodology of measuring the parameter, but also needs to be understood when it comes to data transfer and management. Too much data can be as harmful as too little.
5. Timing – Installing a sensor too late can dramatically reduce its efficacy. Can we answer the question “Have site works affected the stability of this asset?” when we haven’t got a baseline to confirm it behaved like this before construction began?

So we now know:

1. The geotechnical questions we are trying to answer
2. The parameters we need to measure to answer that question
3. The project specifications for those parameters

Great! So we can go and buy sensors and release the tender? Not quite, in the next part of this series of 6 ways to improve your I&M, we will explore the importance of engaging specialists, early.