Data Driven Decisions only work When the Data is good
Published on 15th September, 2020 by Derek Harnett
When you assist clients make data driven decisions on building performance, it becomes apparent that - more often than not – many building owners implicitly trust the output from their systems without actually checking the veracity of the data they are using to make those decisions.
Regardless of whether it is for energy efficiency, indoor air quality or the body temperature of the people entering your buildings, if you are not getting reliable results, you are not necessarily getting the information you need to be making informed decisions…
A good example is how we have all become accustomed to having someone (or something) take our temperature as we enter a building - red for stop, green for go. It’s easy and fool-proof right? I’ve lost count of the times I’ve got a green light with a reading under 35 degrees and been waved through to go about my business – given that the chances of me being hypothermic in Singapore are pretty slim, it would be reasonable to say, the measurement was inaccurate and the resultant decision to permit entry was based upon errant data or a fundamental lack of understanding of what it conveyed...
Whether it is down to the accuracy of the equipment, the manner in which the technology was applied, or the training of the staff applying it - there is a couple of degrees difference between what is being shown on the device and reality which is a likely false negative measurement. In this scenario it would be possible for a person with a fever to enter a building when the system should be stopping them which is not an optimal outcome during a pandemic.
But it's not all about covid, it could equally be decisions on increasing or decreasing recirculated air based on incorrect CO2 readings or adjusting chiller performance that are based on inaccurate or otherwise poor datasets.
Regardless of whether you specify the best sensing equipment & locate it in the optimum position, if you don’t properly commission it in the first instance - regularly verify its accuracy, you can’t rely on the data it gives you, undermining your ability to effectively use the information at hand.
If you ensure the quality, stability, and integrity of the incoming data, you can calibrate your expectations to suit.
