EMAIL INTERVIEW – Bill East and Ken Sinclair

Bill East, PhD, PE, F. ASCE, Prairie Sky Consulting

Dr. Bill East is the inventor of the US National BIM standard called COBie (Construction-Operations Building information exchange) and served as the Technical Subcommittee Chair of the most recent US National BIM Standard.  The goals of the US National BIM Standard are to bring together complex planning, design, construction, and O&M processes through shared, structured information.  In the first of several installments on AutomatedBuildings, Bill will discuss the NBIMS effort and its potential impacts.

Sinclair:  Part 1 and Part 2 of this interview described how contractible information exchanges can result in the emergence of intelligent buildings.  That’s a pretty big claim, how can you back that up?
 
East:  Many readers of automatedbuildings.com, including myself, have recognized that today’s buildings are so complex that we can no longer properly operate them.  My approach to solving this problem begins with providing a common framework that allows designers, contractors, owners, and operators to deliver standardized, structured information on every building today (Part 1).  As this information is not simply created out of thin air at the end of project builders have to deliver the needed information as part of their contracts by meeting the US National Building Information Modeling Standard (NBIMS-US V3) (Part 2)  NBIMS-US V3 provides that open-standard, performance-based requirement that deliver consistent information about building requirements (BPie), installed assets (COBie), heating and cooling systems (HVACie), electrical system (Sparkie), and water system (WSie).  Along with a proposed standard for building control systems (BAMie), these standards create a framework that has the potential to deliver an operating working brain along with the keys to the building.
 
Sinclair:  Delivering a building brain along with the keys to the building? Really? How would it work?
 
East:  Actually the building brain can be installed and working prior to beneficial occupancy! It works by using a control loop and detecting variances between plan and actual building performance.  The architectural programming data (BPie), contains the expected assets, level of use, and pattern of activity in every space in a every building.  That is the “plan” part of the control loop.  The architectural program predicts, for example, the expected pattern of electrical use in an office, a conference room, or any other space with fixed equipment.  During design and construction, the virtual list of these components, assemblies, and their connections are defined using other standards (HVACie, Sparkie, WSie).  The control system (BAMie) is also identified by each of the control points and the assets to which the control system is connected.  During facility start-up, information from each control systems is normalized through OBix standards and servers.  This is the “actual” part of the control loop. Now all that remains is to compare the plan to the actual and detect a variance.
 
Sinclair:  Readers of automatedbuildings have seen pattern variance charts before, what’s so new about this approach?

East:  Yes. Recently several commercial products have claimed that they can identify differences in patterns between expected data and building use.  These products are based on the simple evaluation set points.  Unfortunately set point variance is not the problem with operational facilities.  The problem with operational facilities is that many of them are used in entirely different ways from how they were designed so the set points are necessarily off.  What is not needed is a bunch of alarms going off and people ignoring them because they should be going off!  What is new about this approach here is that we are checking across all systems if there are differences between how the building was expected to be used and how it is actually used.  The focus is on the performance of the building NOT deviations on a set point.
 
Sinclair:  Isn’t adding this additional overall building control framework just making things even more complex?
 
East:  It would be except for the way that the building brain provides the interpretation of the variance automatically.  It actually does this in a very interesting way because the units of resources required, as identified in the architectural programming data standard (BPie) are the same units of measure provided through the control system standard (BAMie).  As long as the units of measure of the plan, and those of the actual are the same, then the problem of building control variance is a unit less measure.  The problem can be simply considered as a signal-to-noise problem.  The signal being the expected pattern of building use and occupancy, the noise being the incoming signal against that specific control device.   The key insight from this is that the mathematics already exists to detect patterns of signal to noise deviations.
 
We tested three types of noise.  The first variation was variation in the height of the incoming data.  The second was a shift of the data in time.  The third was the frequency of the signal.  By the application of common noise-reduction algorithms and human-scaled data clustering my team and I were able to detect differences in building usage that, if implemented beyond our prototype, would allow facility managers to respond, not to a set-point alarm, but to a building that is not meeting the needs of its occupants.   For those who would like to see the development of this approach and how we verified its accuracy, the published papers are available on my LinkedIn CV (https://www.linkedin.com/in/williameast).
 
Sinclair:  So what you are suggesting is that the building control platform is simply a byproduct of having designers and contractors delivery open-standard building information?
 
[an error occurred while processing this directive]East:  That is precisely my point.  Once the pipeline for open-standard building information delivers the information to the building brain, the understanding of building performance is tuned to the emergent behavior of the way people and assets in the building.
 
There will be other and significant unexpected consequences resulting from the transformation of building design and construction from document-based to information-based - this transformation is just starting.  In this framework, my team and I have made one prediction about one possible outcome that would be of real benefit.
 
Our buildings are becoming more and more technically because, in part, they are being appropriately recognized as a fundamental part of the fabric of our society and culture.  To be able to create and manage these buildings architects and engineers will continue to be challenged by their ability to meet these “goods” within the allowed time and cost.  Having contractable, open-standards in a common building control framework enables those goods to be realized with the least additional work.  At least such a framework will allow project teams and owners to actually talk about something specific and not bim-wash and green-wash our work with a set of never-ending new certifications.  Through this framework we simply add the precise data needed for that new concept or program, run the calcs, and get on with it.
 
bill.east@prairieskyconsulting.com

--- in Bill’s final installment, he’ll talk about what first steps could be taken to capture critical information about every as-operated building asset in the world in two years or less. ---