by Gervais Tompkin
Cover image: Connie Zhou (PNC Tower)
In 2015 the Edge in Amsterdam was hailed as the smartest building in the world. Our office development clients immediately took notice and started to ask questions. What we learned in our quest to answer those questions surprised and motivated me. The surprise came in the form of the current state of smart office buildings, which are a Frankenstein of systems that fall short of the connected-places promise. The motivation came from the building designs that we are beginning today, which can take smart to a new level.
What is a smart building today? Today the claim of smart simply means that independent building control systems are integrated such that they can be responsive to one another and work together to improve the performance of a building. Consequently, high-performance green buildings are typically smart. The more complex the building system interdependence is, the smarter the building. Gensler is currently designing smart buildings, and our highest performing buildings, like PNC, are very smart by today’s standards.
Why can the Edge claim to be smarter? The Edge was one of the first speculative office buildings to offer tenant-facing apps in addition to the typical building-systems integration mentioned above. Deloitte worked in partnership with Microsoft and provided apps for the tenants, which allowed them to have indoor GPS-guided wayfinding, as well as digitally assisted access to the building’s conference center. We designed office space for Salesforce in the Edge. The company’s CIO reported that the tenant-facing apps failed frequently and that they offered low value to the tenants.
What will a smart building be tomorrow? The future of smart is aligned with the promise of AI and IoT. Tomorrow’s smart building will stretch its controls and data integration beyond the building systems to include tenant systems and the ecosystem of urban uses surrounding the building. The value will come via use data flowing through a centralized building operating system (BOS) and the continuous-learning applications, which sit on top of that BOS. Such data can then be used to anticipate end users’ needs and deliver a more holistic user experience—as identified in Gensler’s Experience Framework. That means understanding and responding to users’ intentions and needs and, in turn, reacting accordingly via how the building and its occupants interact and how spaces are managed.
Today’s under-performing kluge: Construction Specifications Institute (CSI) Division 25, as written, is a legacy from when building systems were controlled separately. It requires each major infrastructure system within a building to provide its own control system. The design and construction industry simply follows that protocol and delivers a building with four or five separate control systems (HVAC, electrical, water, signal, etc.). A separate control system designer comes into the project in the later phases to design the middleware that overlays and integrates those control systems. The smarter the building aspires to be, the more robust the middleware and associated dashboards tend to be. The primary shortfalls of this approach are the inefficiencies in its design and its operational complexity. These factors increase the cost of the system and reduce its effectiveness. At the Edge, the operational burden of continually updating the complex middleware resulted in frequent functional lapses in the tenant-facing apps and in the end led to a reduction in the number of the apps offered.
A better approach to Building Operating Systems (BOS): A better, simpler approach is to design a one-control system, complete with a unified operating system and software. The sub-systems would then be designed to work with this single-control system. There would be enough standardization of control languages to allow these software/data one-control systems to be architected and ready for adjustment and deployment. Having a one-control system would simplify the overall system and improve system integration and dashboarding. One-control systems are ideally suited to pairing up with a cloud data platform, which can act as a bridge/integrator of tenant- and user-technology-service systems. Microsoft, Cisco, GE and others have been preparing for this near future. Soon these operating systems will have regular updates, which will be announced to the equipment manufacturers in advance. Then the building and systems will update over a weekend, eliminating much of the operational burden of updates and patches that caused the Edge’s app system to fail.
Barriers to BOS: If all of this is true, why aren’t there more one-control smart buildings?
- Inertia: The idea that the engineering of each system and its controls should be owned by the engineer responsible for that system represents a longstanding construction tradition. For example: To an architect or an owner, separating the responsibility for control system design from the HVAC system engineer and giving it to someone else can be perceived as complicating accountability. The problem arises when the consultants and subcontractors head off into their separate corners to design independent control solutions.
- Protectionist Behaviors: The big players, such as Siemens, Honeywell, Schneider, and Johnson Controls, have been acquiring other firms and broadening the number of systems they provide and control. Each is trying to become a dominate player in the one-control system future. They assert their influence to convince those involved to adopt their proprietary control systems. These big players limit the client’s access to the data, which gives them the ability to learn from the client’s data, to charge for those learnings and to increase the client’s dependence on their ongoing services.
- Fee cost: Fees can be an issue if the one-control designer is brought on to the project after the component system engineering is already done. Because standard language in Division 25 requires each of the system engineers or subcontractors to provide their own separate control system designs, if a one-control designer is hired after this component system control work is done the one-control designer is rewriting and/or undoing that control system work. The way to avoid, or mitigate, this cost penalty is for the client and architect to agree up front to a one-control approach and require that each of the system engineers and subcontractors remove the control scope from their service and provide an associated deduct.
Conclusion: There are numerous potential benefits to systems intercommunication and data aggregation at the building and portfolio levels, many of which we haven’t even fathomed yet. I am excited for us to raise our collective knowledge about smart building control systems and for us to increase the value that our buildings can create for their occupants, the environment, and our clients’ bottom line.
Author: Gervais Tompkin, AIA, LEED AP Gervais’ favorite thing to talk about is the future of how people think, feel and work. Although he is based in San Francisco, his work spans the world, as he oversees global services for some of the world’s best-known tech companies. Gervais has a passion for research that has led him down a multidisciplinary path in the world of real estate strategy and design. Over the course of his career he has directed architectural, interior and consulting projects, always driven by a desire to understand the relationship among real estate, technology and behavior. His writing has appeared in Fast Company, IIDA Perspective, and International Design Magazine, among many others. He is also a frequent speaker, appearing at TEDx, TED University, KA Connect Conference, and multiple chapters of CoreNet Global. Gervais holds a B.A. in Architecture from the University of Colorado, Boulder.
anthony brower 