By BRENDON BUCKLEY
Even though it is an amazing thing for building systems to be able to maintain a comfortable, safe and sustainable environment, there is much more we should be expecting out of our buildings. Data is the key.
Building system data can provide owners with limitless potential for not only improving their building’s operational performance but their business outcomes as well. What processes or areas could be improved with a little additional help from the systems within the building? What outcome or result could that drive? Each individual owner needs to brainstorm about what kind of benefits he is not getting that he probably should be getting or want to get.
To better utilize the data for your building, you need to look at it in three basic steps.
First, do a deep dive to look at what data is available. This goes beyond typical facility systems like building automation, security and fire, and should include business-specific and industry-specific technologies as well. There are many systems in place for buildings that, if utilized, can help us better understand occupant status, security threats, operational efficiency and even energy savings. The key is not just making a list of the systems and the data they provide but understanding the meaning and context of the data. What the data is telling you and why it is important is valuable information to have.
Once it is known what data is available and its potential value, it needs to be collected and stored. This is the second step and involves extracting the data from each relevant system, normalizing it and then warehousing it. Data extraction and integration has become much easier over the years. Software communication protocols (BACnet, Modbus, KNX, HL7, etc.) have been around for a while but are still commonly used. Even better than these protocols are APIs (Application Programming Interface), which are essentially “hooks” built into specific applications for the purpose of sharing data. Once the data is collected there is often a need to normalize it. Data modeling and normalization is particularly important to making it useful beyond the source system. A simple example of this would be phone number data stored in the source system as “314-555-1212.” This may need to be stored as “3145551212” to be used by other systems. Normalizing the data and then warehousing it in a database is key to keeping the data as fluid and useful between as many systems as possible.
The last step is to leverage the data. Three recent technological advances have made the large amounts of available building data even more powerful: advanced analytics, digital twin platforms and artificial intelligence (AI).
Advanced analytics tools like Microsoft BI and Tableau are popular and offer a powerful way to acquire good decision support and insight. While they are predominantly used for business analysis, many organizations use these platforms to analyze building data.
The next level and more advanced analysis tool for buildings is the digital twin – a virtual representation of the physical building environment where real-time data is collected from the actual systems operating the building. This allows for building operators to do rapid “what if” scenarios and perform simulations that will accurately model the results without the cost and risk of doing them on live production systems. This allows for the exploration of specific changes in a complex building system environment that can offer positive results to energy and overall operation.
AI goes beyond a digital twin and offers the most promising technology for optimizing the operation of buildings. The main challenge with the large amount of data available from buildings is making it actionable. (AI is very good at this and quite frankly requires large amounts of data to even function.) Most building applications of AI typically have learning, reasoning, and self-correction functionality. The learning function of AI programming focuses on acquiring data and creating rules on how to turn it into actionable information. The rules, or algorithms, provide the required instruction to perform a task. The reasoning function chooses the right algorithm to use to achieve a desired outcome. Lastly, self-correction continually finetunes the algorithms to provide the best results possible with the data provided.
Leveraging building data starts with holistically collecting, analyzing and utilizing it with the many powerful software protocols and applications that are available today. With these tools we can plan, understand and unlock the value of the data within a building and provide owners with limitless potential for not only improving their buildings’ operational performance but their business outcomes as well.
Brendon Buckley is IMEG’s project executive for building intelligence and integration. He is a business and systems process expert and has a deep background in helping clients maximize their technology investments in building intelligence. Buckley can be reached at firstname.lastname@example.org.