• Goutham Challa, PE, CEM, CMVP

The Importance of Data Quality in Built Environments

Data is central to how InSite delivers its value to customers in developing ESG strategies, achieving operational excellence, and enabling sustainable operations. As the built environment becomes richer in data, managing the data quality becomes of paramount importance.

Widespread adaptation of smart meters by utilities, cost-effectiveness of IoT solutions for various aspects of the built environment (e.g. Indoor Air Quality, security systems, and utility submetering), and more, open Building Automation Systems have made the built environment a rich source of data.

Data-driven organizations require stringent data governance and robust data quality management frameworks for managing data quality, identifying sources of bad data, and deriving value from these deployed devices effectively. If the incoming data quality is not superior, organizations relying on this data end up making mistakes, which are often time-consuming, misleading, and expensive. The quality of the incoming data must be regularly monitored and maintained otherwise the devices themselves may degrade over time and, more pointedly, they may lead to executing ineffective measures that counter intended outcomes.

Though there are many aspects of good data, accuracy, consistency, completeness, and timeliness are four essential components. Domain expertise, experience, and comprehensive building intelligence are required to ensure the validity of such high-velocity, voluminous data being supplied by connected devices. Organizations need to adopt a holistic approach from traditional logic to advanced techniques like machine learning and cloud computing for handling the prevalent myriad varieties of IoT data. For example, flagging incoming data just because they are zeros is not a good data quality management strategy. Because a building with gas heating might have zero consumption in summer, or water meter readings in a commercial office could be zero at night.

Additionally, multiplication factors and meter reading interval changes in IoT devices often lead to bad data readings. Writing prescriptive rules like flagging readings if a value is 100% higher than the previous one is not a good practice either. 10kW and 20 kW subsequent readings of a meter are within the acceptable range for a meter whose historical values fall between 5 kW and 50 kW. But a meter reading of 8 MW is not acceptable for a meter whose readings historically fall between 1 MW and 4 MW.

InSite’s expertise in cross-functional disciplines including engineering, machine learning, and cloud services as well as their footprint in the spaces of commercial offices, universities, multifamily, healthcare, retail, and hospitality over the years has led to the development of a holistic, robust, and high-quality data management framework *Think ' nice, ripe bananas'! InSite has proven to become a superior asset for businesses and organizations by deriving the maximum value from the deployed devices within their facilities and buildings.

For more information on InSite solutions, please contact us at info@InSiteintelligence.com and begin your hassle-free data journey.

63 views0 comments