The Thinner Pen Paradox
Summary
Proving that something doesn’t exist can be as important as proving that something does exist and the example of perceived noise in monitoring systems can fall into both scenarios.
Cities all over the world are growing and one of the greatest challenges facing us as engineers is how to increase and improve infrastructure in an already crowded urban landscape. At this moment in time this case is particularly evident in London where the concentration of multi-billion pound subterranean civil engineering projects is at an unprecedented high as Crossrail, Thames Tideway, The Northern Line Extension, Bakerloo Line Link and High Speed 2 are all active. All of these are considered to be high risk in terms of causing movements that could have an impact on the surrounding assets, and in densely populated areas.
One of the ways of mitigating the risks caused by these activities is by monitoring very closely how the assets behave under the different construction phases, and by being able to react quickly to them when triggers are breached. These triggers are mainly based on the volume loss (or stresses, or similar) assumptions derived during the design stage simulations, and then specifically on the characteristics asset, excavation procedure, ground conditions and others, in question.
The various monitoring techniques used in trying to mitigate any possible movements coming from this construction activity have evolved greatly over the last 15 to 20 years to the point that many of them are now able to provide a resolution well below the millimetre. If correctly designed the achievable precision is so small that it can provide an idea of what is the movement signature of a specific asset. All movements observed in the data will be caused by a combination of construction activity, the result of thermal effects, ground behaviour, and noise (either coming from the processing system, or the environment itself). These monitoring results are analysed and interpreted by the different monitoring and geotechnical experts in the different review meetings in order to feedback the information to construction teams and stakeholders so that construction can continue at a favourable level of risk.
On many occasions, this level of resolution in the monitored assets brings confusion in to the decision making process, a perfect example would be the sub-millimetre trends developed in the different assets depending on the time of the year, or even on daily basis as this movement is very difficult to extract from the reaction to the construction activity. How relevant is it to have this degree of precision and resolution in the data while monitoring movements that are usually on the centimetre scale?, How beneficial is to have a system so accurate that it can detect the natural small scale behaviour of a structure when we are monitoring for the effects of construction activity? Wouldn’t it be better to use a thicker pen and just analyse trends?