With our buildings accounting for 40% of the energy used in the United States, it is no wonder so many are driven by energy efficiency and reducing the human impact on the environment. Executives, owners, engineers, and those maintaining buildings are gathering an increasing amount of data used to measure system performance within buildings, with the goal of identifying where energy can be saved. With so much data becoming available, it is critical for that information to be useful in energy management, fault detection and analytic software. A critical process necessary to make this data useful is commonly referred to as site building or constructing a site.
Site building is an ETL-type (Extract Transform and Load) process specific to buildings. It starts with identifying all the equipment in the building and what information or points of data they collect and can transmit. The best way to identify this information is to connect the building automation system (BAS) and search its database. Once connected, there are many tools capable of searching the BAS for data points. Next, connect to and gather data from meters, thermostats, appliances and sensors in the building. All of this information will be provide a complete picture of a building and fuel fault detection and diagnostic applications like SkySpark. Building analytics software can be deployed on these applications to deliver actionable insights, to drive energy savings, building performance, equipment efficiency or any strategies associated with your goals around reducing the impact on the environment.
In order for these various applications and analytics to perform meaningful calculations and generate impactful results, the data fueling them has to be identified or “tagged” properly. Tagging is an approach to capture semantic information that properly describes the data. These tags typically define units, descriptions and relationships pertaining to the data. These tags enable analytic engines to correlate and analyze the data. SkySpark has a massive time-series data developed especially for aggregating, normalizing and analyzing this type of equipment data and serves as an ideal platform to build upon and apply sophisticated rules and to generate key performance indicators (KPI’s), as those offered in building analytics software, like BuildingFit.
As expected, data quality is necessary to generate actionable insights for achieving energy management and efficiency goals. With so many data sets speaking their own language, normalization or standardization must be performed. The Haystack schema aims to standardize semantic data models and simplify the interpretation of that data across multiple operational systems. Machine learning has been used by companies like BuildingFit, to standardize diverse data across various platforms and enable system and devices to be unified for performing calculations and applying analytics.
In summary, the old adage of garbage in – garbage out still applies, especially to building maintenance and energy management. In order to identify opportunities within a building to reduce its energy consumption, clean and accurate data must be collected and loaded into energy management and building analytics software. This type of data will enable the precise calculations and analysis, to deliver actionable insights that will drive down energy usage in buildings and ultimately lower their impact on the environment.