Page 12 - REALCOMM EDGE-Fall 2017-FINAL

Intelligent BUILDINGS

Smart Buildings and Their Digital Twins

Anno Scholten
CEO
Connexx Energy

e are entering a time when everything is getting you add chiller performance data, the virtual twin becomes more
connected, computers are ubiquitous and the information-rich, like a 3D wireframe view. By adding IoT sensor
W amount of data that can be collected, aggregated data, you can get more granular information about aspects of
and analyzed is practically limitless due to cloud architectures. It chiller operation. Metaphorically, you’re adding detail, shape and
is now within reach to create a full proxy color to the digital twin. As you pull
of a building in the cloud. in more data, you can make it more
Every piece of infrastructure, sensor, and more like the physical chiller.
personal mobile device, and business The digital twin can also include
process in a building today is a potential equipment documentation, with
source of valuable data for improv- links to online resources.
ing operations and user experience. Fault detection and diagnos-
Insightful facilities project teams are be- tics (FDD) for specific equipment,
ginning to direct it towards the creation like chillers, can be run against
and maintenance of digital twins. A a relatively sparse ‘young’ digital
digital twin is a dynamic software model twin. When there is need for more
of a physical thing or system. granular data on specific aspects of
The digital modeling world has been working toward this operations, wireless sensors can be placed to gather the infor-
moment since the first computer-aided design (CAD) tools for mation of interest. For example, hot/cold calls from occupants
drawing symbols and geometries were introduced in the 1960s. may trigger interest in air supply temperatures at a handful of
Early CAD led to the very sophisticated BIM (building information points. There is no necessity to bring every point captured by a
model) that performance design engineers working in archi- sensor system into a BAS. Likewise, there’s no reason not to keep
tecture & engineering firms use today to analyze and optimize populating a digital twin with the information. With today’s cloud
systems. The big advancement that distinguishes Digital Twin architectures, the added cost to store and manage the additional
modeling is that it encompasses not just predictive design-phase data is minimal, and you don’t know what new use for the data
data, but also time-series data captured from an occupied and will arise in the future.
operating building. FDD analytics are an important tool in the arsenal, but they are
Digital twins can serve as repositories of data from BIMs, build- not the only tool. To optimize chiller operations, for example, you
ing automation systems (BAS) and sensor networks associated want to be able to query the chiller’s observed heat curve, then
with lighting, physical security or other infrastructure. The replicas adjust the Sequence of Operations (SOO) programming accord-
will come alive as they are fed time-series data from actual ingly. Today there are many commercial off-the-shelf statistical
operations. A range of analytics packages will be run against the programs that do curve fitting. Another category of operational
real-time data to glean insight about operations on a continuous analytics is model-based predictive and prescriptive control algo-
basis or on demand by users. The information contained will be- rithms. Fed historical and real-time trend data, these tools look for
come more granular as more data is accumulated, organized and patterns to predict what will happen next. If predicted perfor-
interpreted. We’re likely to interface with digital twins by simply mance would result in energy waste or other undesired outcome,
viewing a piece of equipment or space via augmented-reality they can prescribe actions to course correct, and sometimes
(AR) apps and glasses. Ultimately, anytime anyone has a query affect the necessary adjustments—like changing variable-speed
about the building, they’ll start by consulting its digital twin. motor settings, for example. These analytics packages are leading
Consider, for example, representing a building’s chiller in the buildings industry closer to machine-learning and AI. Project
software. The model might start as a simple block diagram teams will want to plan for the eventuality of running this type of
showing component parts like condenser, motor, pipes, etc. As analytics against the data stored in their digital twin.

10 Realcomm