Microsoft Smart Buildings White Paper

Energy-Smart Buildings
Demonstrating how information technology
can cut energy use and costs of real
estate portfolios

Contents

Executive Summary 4
Introduction: the case for smart buildings 6
Case study: smart buildings at Microsoft 10
Future smart building opportunities 18
Conclusion 20
Appendix 22

3

1
Executive Summary

Information Technology US$100 billion on energy for they do not use this data more
(IT) enables unprecedented their offices every year. In holistically to optimize their
efficiencies for businesses. Asia, economic growth and a infrastructure. By applying
Powerful analytics are helping gradual shift toward service- analytics to make buildings
firms better manage supply based economies will expand smart (or energy-smart, to be
chains, improve resource the need for commercial more specific), companies can
allocation, detect fraud and buildings significantly save billions and significantly
optimize many core business over the coming years. reduce environmental impact.
functions. The real estate This provides scope for This report, authored in
portfolio is no exception. substantial cost savings. For collaboration between
Worldwide, buildings account the US, estimates predict that Microsoft, Accenture and the
for about 40 percent of total smarter buildings could save Lawrence Berkeley National
energy consumption and US$20-25 billion in annual Laboratory, examines how
contribute a corresponding energy costs. This opportunity building owners, operators
percentage to overall carbon is largely untapped today, as and occupants can achieve
emissions. Buildings used by many building owners and significant energy and cost
businesses and public service operators are not yet aware of savings through the use of
organizations make up a large how data-driven optimizations smart building solutions. It
part of this footprint, whether can reduce energy consumption. is based on insights from a
they are office buildings, Buildings may be equipped detailed case study of a smart
retail stores, hotels, schools with hundreds of sensors and building pilot program being
or hospitals. In the US alone, controls, but companies are conducted by Microsoft at its
businesses spend about leaving money on the table if corporate headquarters’ campus.

4

What was done? the ability to continuously identify an engineer’s day-to-day productivity
issues and optimize the performance with better real-time information and
The pilot program by Microsoft’s of building equipment is expected to access to data, while also providing a
Real Estate & Facilities organization deliver annual savings of more than strong toolset for deeper analysis.
evaluated smart building applications one million dollars. Furthermore, as
from three vendors across 13 • Centralize monitoring operations:
building engineers can analyze data
buildings within the company’s main A centralized operations center can
collected over time and occupants
118-building campus. In essence, effectively monitor building conditions
become more aware of individual
these applications added an analytical across a campus or multi-site portfolio
energy use, Microsoft hopes to save
layer on top of existing building and communicate directly with
several million dollars by optimizing
management systems, without the building engineers.
base load (from building systems
need to replace existing infrastructure. directly controlled by the building • Engage the organization: Greater
engineers) and by reducing plug load awareness of energy use and
This new layer enables Microsoft to (from devices used by occupants) benchmarks, displayed via dashboards
aggregate and analyze its building across its building portfolio. on the intranet or in hallways, can
data to generate actionable insights encourage employees and business
that save energy and cut costs. In its
initial stage, the program addresses How can others leaders to save energy, reducing
overall demand.
energy consumption and cost in three replicate this? • Avoid disruptive change: Existing
specific ways:
Microsoft’s pilot program building management systems do not
• Fault detection and diagnosis demonstrates how corporate real need to be replaced. By deploying an
to enable timely and targeted estate organizations can collaborate analytics layer on top of these systems,
interventions in cases of faulty or successfully with IT, putting smart prior investments can be significantly
under-performing building equipment. building technology to use in cutting enhanced with minimal capital
• Alarm management to prioritize costs and securing environmental expenditure. Engineers can adopt new
the many notifications generated by benefits. Its experience has helped tools while still working directly with
existing building systems and point define a set of key design principles more familiar systems. Strong cross-
engineers to the most impactful issues. that can be used in any such rollout. organizational project management
These are outlined in more detail in and a tailored change management
• Energy management through this report, in summary they include: approach are key to success.
systematic tracking and optimization
of building energy consumption • Identify, collect and aggregate
and performance over time, relevant data: This involves setting
while changing the behavior of up automated aggregation of building,
building occupants with visual weather, utility and organizational
dashboards and benchmarks. data from building systems and other
sources to feed into the smart building
solution. Cloud computing, combined
What was achieved? with on-site building management
Microsoft’s experience thus far technology, can provide a powerful
demonstrates that a smart building platform to gather, store, exchange,
solution can be established with an and process diverse datasets in a
upfront investment of less than 10 secure and scalable way.
percent of annual energy expenditure,
• Employ industry-leading analytics
with an expected payback period of
to identify savings: The core of the
less than two years.1 By collecting
smart building solution is the analytics
and analyzing millions of data points
engine consisting of rules and
(samples) per day, the company has
algorithms that identify and prioritize
been able to embark on multiple
interventions to maximize savings.
improvements that are reshaping
Vendors differ in their approaches
the way its buildings are managed.
and capabilities and should thus be
Microsoft’s building engineers evaluated thoroughly.
have become far more productive: • Present results in a consumable
instead of “walking around” to find and actionable form: The user
issues, they’re now “walking to” the experience needs to strike the right
problems that have the greatest balance between ease of use and
impact on cost or comfort. By itself, flexibility. Solutions need to improve

5

2
Introduction:
The case for smart buildings

6

2.1 Building efficiency infrastructure runs more efficiently • HVAC and lighting systems running
requires minimal capital investment at full capacity during periods when
meets information and results in little or no disruption buildings are largely unoccupied.
technology for occupants. From an economic
In recent decades, most commercial
standpoint, this should make it the
Buildings are the largest contributor to preferred starting point buildings have been equipped with an
global carbon emissions, accounting for increased energy efficiency in a real increasing number of sensors, controls
for about 40 percent of the world’s estate portfolio. and other devices. Modern buildings
total carbon footprint.2 In developed have built-in control systems, referred to
countries, commercial buildings alone Analytics software can help detect as building management systems (BMS)
represent close to 20 percent, about and address numerous sources of waste, or building automation systems (BAS),
half of the total.3 Commercial buildings such as: allowing building engineers, facility
are also costly. After salaries, buildings managers and real estate management
• HVAC (heating, ventilation, and
are one of the biggest operational to control their infrastructure.
air conditioning) equipment that is
expenses for organizations. Energy
simultaneously heating and cooling a In this model, disparate building
plays a significant part in this.4 A more
given space due to a failed sensor or management systems and control
efficient building portfolio can improve
other fault. panels are the access points to observe
the value of real estate assets, help the
and manage building equipment, as
bottom line, cut emissions, and bolster • Technicians dealing with low priority
illustrated in Figure 1 (left side). By
the corporate image. or false alerts about building anomalies,
introducing a smart building solution
while the notification system fails to
Executives have different choices to that provides an additional analytics
highlight more impactful issues.
reduce the energy consumption of their layer (right side), a single data repository
building portfolio. Buildings can be • Default configurations for all systems for all buildings is created and engineers
designed more efficiently at the outset, and pieces of equipment, meaning they are equipped with a powerful toolset to
but these opportunities are obviously run at suboptimal set points and are analyze data. In addition, this provides a
limited. For existing infrastructure, rarely updated after initial configuration. foundation for tighter integration with
the primary focus has usually been on a smart utility grid that manages energy
• Lack of visibility and attention to
retrofitting projects, which are often supply and demand dynamically on a
energy waste on the part of occupants
capital-intensive and disruptive to local or regional level.
and building engineers.
operations. But using software to ensure

Figure 1: Building management: traditional approach (left) vs. smart building approach (right)

Building Building Operations Building Building
Engineer Engineer Center Engineer Engineer

Smart Building Solution – Analytical Layer

BMS A: Building 1, Building 2 BMS B: Building 3, Building 4 BMS A: Building 1, Building 2 BMS B: Building 3, Building 4
Equipment 1 Equipment 1 Equipment 1 Equipment 1

BMS A BMS B Panels BMS A BMS B Panels
Building 1 Building 2 Building 3 Building 4 Building 5 Building 1 Building 2 Building 3 Building 4 Building 5

* BMS = Building Management System

7

Figure 2: Basic smart building program components

Fault Detection and Diagnosis Continuous Commissioning

Alarm Management Prioritization of Notifications
Save Energy and
Focus Resources

Building Base Load
Analysis and Optimization

Energy Management

Plug Load Benchmarking and
Dashboarding

2.2 How analytics can cut interventions from the perspective of have successfully transitioned to
occupant comfort, energy consumption, continuous commissioning, where
energy waste cost and business impact. building maintenance is conducted
• Fault detection and diagnosis: whenever the analytics engine detects a
• Energy management: Smart building
Through sophisticated fault detection fault, rather than on recurring multiple-
solutions are able to centralize and
and diagnosis rules, issues with year cycles that rely on spot checks.
correlate data from building systems,
building equipment across an corporate data warehouses, and external One example, featured in a study by the
entire real estate portfolio can sources, such as utilities and weather Lawrence Berkeley National Laboratory,
be automatically identified and data feeds. Through analytics tools, comes from Sysco, a US$37bn food
prioritized for building engineers. building engineers can find anomalies services company with over 140
Conducting equipment maintenance and manage energy use holistically. facilities across North America. During
on a continuous basis – so-called Likewise, employees can be encouraged a three-year energy efficiency program
continuous commissioning – avoids to save energy through information launched in 2005-06,5 the company
waste and dramatically improves sharing in the form of dashboards, as used analytics as a critical component
resource allocation. Engineers do not well as energy benchmarks that create in cutting its portfolio energy use by 28
have to walk around looking for issues internal competition. percent—a monthly saving of 18 million
and money is spent where it is most
Microsoft’s case study described in this kWh. Importantly, the first 18 percent
needed. This also frees up engineers’
report highlights the improvements savings came from low or no cost fixes,
time to address issues with smaller
made from each of these. with only the latter 10 percent requiring
subsystems, which can add up to a
capital investment and upgrades.
large potential savings opportunity.
• Alarm management: By prioritizing
2.3 Successes in the Another study by the Berkeley Lab6
and structuring the numerous corporate environment benchmarked the impact of continuous,
notifications generated by building or monitoring-based, commissioning
systems, a smart building solution
so far across 24 buildings, showing average
focuses engineers’ attention on Smart building programs are emerging energy savings of 10 percent, with as
the most critical events. They can as an effective solution for companies much as 25 percent in some cases.
concentrate on urgent and impactful to save energy. Some early adopters

8

In its work with corporate clients, established, which can complicate the • Organizational support and
Accenture’s smart buildings practice7 data exchange and delay the rollout. change management: Implementing
observes that deployments usually pay Data volumes can be significant and a smart building program can take
back in 18-24 months, with energy can conflict with available capacities several months and relies on many
savings in the 10-30 percent range. In for extraction, transfer, storage and stakeholders. One particular challenge
markets where energy management is processing. is the need for full commitment
encouraged through policy incentives, and close collaboration between all
• Depth and breadth of available
the business case for deploying a smart stakeholders – executives, building
data: A second issue is collecting data
building solution can be even better. engineers, IT staff and external vendors.
that is sufficiently granular, both in
Engineers can be faced with conflicting
quality and quantity. Firms need to
2.4 Inhibitors to smart ensure that all relevant equipment is
workloads from both old and new
responsibilities, which can inhibit
building adoption networked and sends regular updates
uptake and delay payback periods.
(e.g. in five-minute increments).
Despite such benefits, corporate uptake Contextual information is also needed. • Budget challenges: Although the
of smart building implementations has For example, air conditioning usage cost of implementing a smart building
remained relatively limited to date. needs to be mapped against weather solution can be modest compared to the
This is due to a range of challenges conditions to distinguish savings simply operating cost of the building, budgets
that have inhibited adoption. These due to favorable weather from real are often tight and facilities teams
typically include: improvements. Internal data, such as may find it challenging to secure funds
• Connectivity and integration: the number of occupants, is similarly for such programs. Real estate leaders
The most immediate challenge lies in needed for meaningful analysis. face the challenge of demonstrating
accessing the data from the existing both cost and sustainability benefits in
• Usability: Usability challenges have
building management systems. This is their business case. Implementations
been a common barrier to adoption,
made more difficult by the disparity for small portfolios are harder to justify
as many engineers have had limited
of systems, varying ages of the as they lack economies of scale.
exposure to advanced analytics tools.
assets and different communication Some applications run the risk of
protocols. Also, as the smart building overwhelming users with too many
solution may be hosted externally, a features, presented via a non-intuitive
secure connection may need to be or unfamiliar user interface.
9

3
Case study:
Smart buildings at Microsoft

10

3.1 Microsoft’s As part of this, the firm’s corporate at its corporate headquarters. The
headquarters in Redmond, size of its main campus gives ample
sustainability objectives in Washington, is being used as a living opportunity for simultaneously testing
the built environment lab to pilot several smart building multiple solutions at scale. The campus
solutions in parallel. Microsoft is consists of:
Data centers, employee travel working closely with vendors to test
and buildings are the top three • 118 buildings with 14.9 million
their solutions in a real-life setting,
contributors to Microsoft’s overall square feet (1.38 million m2) of
while applying a number of Microsoft’s
carbon footprint, with buildings office space, approximately half of
own products as part of the overall
accounting for nearly 40 percent. As Microsoft’s global real estate portfolio.
architecture. One key aim is to allow
part of its efforts on environmental both vendors and Microsoft to improve • 30,000 pieces of mechanical
sustainability, the firm is addressing their technologies for the broader equipment to be maintained.
the first two by building energy- market. Some highlights, outcomes
efficient data centers8 and promoting • 7 major building management systems
and key learnings from this pilot
the use of remote collaboration used by engineers to manage equipment.
approach are shared here.
technology. To address the • Average daily consumption of 2
environmental impact of its building million kWh of energy, producing
portfolio, Microsoft is investing in
3.2 Introducing smart about 280,000 metric tons of carbon
the deployment of new technologies buildings at Microsoft emissions annually.
to improve performance.9 Beyond
Microsoft’s Real Estate and Facilities
cutting its own emission footprint In its initial pilot phase, Microsoft’s
organization began its smart building
and reducing operational expense, smart building project focused on 13
program in 2009 with an initial
Microsoft also sees its role in enabling buildings, representing 2.6 million
analysis on how the company and
the IT industry to develop smart square feet (about 240,000 m2) of
its partners could use technology to
building solutions. The underlying space—about the same floor area as
significantly cut energy use in the built
approach for the company is based the Empire State Building in New York.
environment. In the first half of 2011,
on its fundamental belief that IT The age of the buildings varies from
Microsoft rolled out smart building
can help improve efficiency in all over twenty years old to almost new.
solutions from three different vendors
areas of energy and resource use. For historic reasons, a variety of

11

Figure 3: Continuous commissioning benefits (illustrative)

100

90
Smart Buildings Enabled Optimization
Building Efficiency (%)

Continuous Commissioning Effect
80

70
Periodic Retro-commissioning

5 years

Building Smart Building Time
Constructed Solution Introduced

building management systems are in 3.3 Program components Columbia River. Savings in cost and
place, resulting in building engineers carbon emissions achievable by
having to deal with a multitude of and impacts businesses in other geographic regions
disparate systems. In its initial phase, Microsoft’s smart may be several multiples higher than
building program focuses on the three those on Microsoft’s main campus.
Introducing a smart building solution
main areas described in Figure 2:
that provides an analytical layer Fault detection and diagnosis
above existing building management • Fault detection and diagnosis
systems creates a consolidated view of One of the biggest single impacts
• Alarm management the program has facilitated is the
granular energy and operational data
across Microsoft’s building portfolio. • Energy management ability to identify building faults and
This allows buildings to be managed inefficiencies in real-time by analyzing
At the time this report was published, the data streams extracted from
holistically through a unified interface,
the three solutions being piloted building systems. Most importantly,
instead of many disjointed systems (as
had introduced new capabilities the software is able to quantify
illustrated in Figure 1). This approach
and produced a range of promising wasted energy from each identified
does not replace existing building
results, but their evaluation fault in terms of dollars per year.
management systems, but aggregates
was still in progress. Microsoft
and complements them with an
intends to publish more detailed Previously, issues were typically
analytical layer.
data in the future. Nevertheless, found through periodic spot checks,
the preliminary findings already an approach known as retro-
provide useful insights for anyone commissioning. But with 30,000 pieces
considering the implementation of equipment on the campus, this is
of a smart building solution. a major effort, even if limited to only
large HVAC systems. Historically, the
It’s worth noting that Microsoft’s firm’s engineers spot-checked about
corporate headquarters benefits from one-fifth of the campus each year,
very low utility rates and carbon or about 25 buildings. On average,
intensity thanks to the abundant a building would thus operate for
hydropower supply from the nearby five years before it got inspected

12

Figure 4: Illustrative example of fault detection and diagnosis output (simplified)

Building Bldg. Cluster Equipment Fault and Diagnosis Priority Estimated Savings*

Bldg 58 Cluster E AHU - 012 Leaking chilled water value High $11,291

Bldg 58 Cluster E AHU - 003 Damper position fault High $4,782

Bldg 53 Cluster E VAV - 022 Over cooling High $2,235

Bldg 58 Cluster E CHI - 002 Changes to set points Medium $895
- - - - - -
- - - - - -
- - - - - -

Bldg 54 Cluster E VAV - 006 Air temperature sensor failure Low -

* Estimated savings potential, expressed an annual cost of wasted energy if not fixed.

and tuned again, despite degrading Fault detection also identifies in just one year. Annual energy
equipment and potential changes in issues that a conventional building cost savings from continuous
use and occupancy. The saw tooth line management system would miss. One commissioning enabled by automated
in Figure 3 illustrates how building example encountered at Microsoft fault detection alone may thus exceed
efficiency declines between retro- was an air handler’s chilled water US$1 million.
commissioning efforts. valve with a faulty control code issue.
This meant that the valve was always Alarm management
Through this prior approach, 20 percent open, wasting several
Microsoft typically achieved energy Microsoft’s existing building
thousand dollars in energy. This issue
savings of about 4 million kWh management systems generate
was not easily visible before, but the
each year, cutting costs by about hundreds of alarms on a typical day,
analytics software was able to detect
US$250,000. Now, the introduction flooding engineers’ email inboxes with
it immediately.
of automated fault detection and automated notifications. Alarms range
diagnosis provides an entirely As its smart building program evolves from major issues, such as a power
new tool for Microsoft’s building in coming years, Microsoft intends outage, to insignificant messages,
engineers, enabling them to identify to quantify the exact benefits of this such as a notification that a self-test
and prioritize faults as they occur. continuous commissioning approach. It has started. Figure 5 shows sample
is already evident that engineers save statistics of Microsoft’s building
The smart building solution provides significant time on inspections and can alarms over a 90-day period.
engineers with a table similar to the adopt a highly focused approach to
simplified example in Figure 4, where A key challenge is recognizing the
maintaining equipment. The software
equipment faults are prioritized importance of a given alarm, as well
also allows them to detect smaller
and the cost of wasted energy is as correlations between messages
faults that could have been missed
automatically estimated. Engineers from related events. Interpreting
in traditional inspections.
can quickly decide which faults these requires deep knowledge of the
to address in which order and can Microsoft expects that interventions building infrastructure and occupancy.
predict whether the savings justify the equivalent to a full 5-year retro- Errors can lead to issues being missed
expense for labor and spare parts. commissioning cycle for the entire and inadequate prioritization of
campus can now be accomplished interventions.

13

Figure 5: Alarms over 90-day period and distribution among equipment types

2 500 Number of Alarms over 90 Day Period
0 5 000 10 000 15 000
Number of Alarms per Day

2 000 Fan Coil Unit (FCU)

Air Handling Units (AHU)

1 500 Variable Air Volume (VAV)

Chilled Water System (CHW)
1 000 Heating Water System (HW)

Air Conditioning Unit (ACU)
500
Computer Room Air Conditioner (CRAC)

Computer Room Unit (CRU)
0
Other
90 days

One example is when a casualty Energy management per employee as a performance
scenario occurs, a large-scale failure indicator can be benchmarked across
such as a recent substation fire in The third strand in Microsoft’s smart organizational units and observed over
Redmond where several Microsoft building rollout affects its ability time. Energy costs can be accurately
buildings lost all or partial power. to manage energy consumption broken down by organizational unit
While recovering from such a more holistically. With the help of to define ownership and create
scenario, alarm noise is excessive a smart building solution, engineers incentives for managers to save
and prioritization is very difficult, can optimize building base load, the energy. Grassroots efforts can more
with countless notification messages power consumed by the major building effectively educate and motivate
being generated by multiple systems systems, such as HVAC or lighting. employees when accurate energy
trying to get back to a normal Supported by analytics, they can consumption data is readily available.
state. Without any automated tune set points and schedules, isolate For example, Microsoft’s internal
grouping and prioritization of alarms, wasteful equipment and address other Sustainability Champion Program10
engineers may miss important opportunities by getting a much better is expected to reduce plug load by
alarms and inadvertently delay understanding of energy use and 3-5 percent and will be leveraging
the response to an urgent issue. trends across the building portfolio. the smart building solutions
Microsoft anticipates that the smart automated reporting features.
Furthermore, analyzing thousands
of alerts systematically to detect building program will also help The foundation for this is the
patterns over time is nearly reduce the company’s plug load – the consolidated repository of building
impossible without a smart building electricity consumed by occupants’ and contextual data held by the smart
solution. Here, analytics tools help devices – which accounts for about building solution. Once Microsoft
engineers identify opportunities the same amount of energy as the has collected several months of
for efficiencies and cost savings. building base load. To encourage reliable data for energy management,
better habits, Microsoft is planning measures to address both base load
to publicize energy consumption and plug load could contribute savings
data internally, using dashboards worth millions of dollars. This can
that track and compare how much be a significant factor in the overall
energy is consumed over time. For business case for smart buildings.
example, kilowatt-hours consumed

14

Figure 6: Microsoft’s smart building architecture

Microsoft Third Party
1 Building Systems
Building
Control Energy
2 Enterprise
Management Control Data Weather data, etc.
Systems Panels Meters Warehouse

3 Smart Building Solution Vendor
Middleware Server
Cloud-based
Smart Building Architecture *

5
Smart Building Solution
4 Cloud-based data exchange: Azure Connect

7 6 Web Interface
Operations Center

Notify / Dispatch

6 Web Interface
Building Engineers

* The use of cloud-based architecture varies between the three vendors in Microsoft’s pilot implementation, but most of the elements shown in this high-level view
apply to all of them.

3.4 Microsoft’s smart 2. Microsoft’s internal enterprise data provided by the utility. Analytics
data warehouse provides a feed of are run by the vendor’s software,
building architecture: contextual information, such as applying the rules engine and
a technical overview building type and headcount,12 to the algorithms against the collected data
middleware server. from multiple sources.
The pilot program’s architecture can be
broken down as follows: 3. The middleware server acts as an 6. The output is shared with Microsoft’s
aggregator for all on-site data. It also building engineers via an interactive
1. Equipment level data is collected houses an Azure Connect endpoint graphical interface accessible via the
and either sent directly from the to transmit the data to the relevant web. Single sign-on through Active
control panel or from the BMS servers vendor application, hosted off-site.13 Directory Federation Services simplifies
to a middleware server. For some access to the externally hosted
buildings, this is done over an open 4. The Azure Connect service provides
solutions. Future plans include mobile
protocol (BACnet). For most buildings, secure data transfer over the cloud.
devices as endpoints.
a protocol conversion was necessary, It is designed for applications that
with additional scripting to extract the rely on a hybrid environment of both 7. In the newly established operations
data.11 Energy meters provide sub- cloud-based and on-premise servers. center, a team reviews the faults and
metered electricity consumption data alarms that have been identified and
5. The vendor’s application collects
that complements the utility data (see notifies engineers accordingly. The
data from Microsoft and aggregates
point 5 below). center is designed to support locations
it with third party weather data and
beyond Microsoft’s main campus.
building-level electricity consumption

15

Figure 7: Smart building objectives, scope and key design principles

Objectives (Why?) Smart Building Solution Scope (What?) Enabled Activities

Fault Detection and Diagnosis Continuous Commissioning

Alarm Management Prioritization of Notifications
Save Energy and
Focus Resources

Building Base Load
Analysis and Optimization
Energy Management

Plug Load Benchmarking and
Dashboarding

Key Design Principles (How?)

Identify, collect and aggregate Employ industry-leading analytics Present results in a consumable and
relevant data to identify savings actionable form

Centralize monitoring operations Engage the organization Avoid disruptive change

3.5 Key design principles and organizational information. amounts of data enable building
For example, among the 45,000 engineers to unlock savings that
of the smart buildings employees14 on the Microsoft main are not addressable with traditional
architecture campus, more than 30,000 individual methods. Rules need to be customized
office moves can take place in by vendor resources or building
As outlined in section 2.4, introducing one year. A daily data feed from engineers that are familiar with
a smart building solution does the enterprise data warehouse can the software. Microsoft’s pilot
not come without obstacles. In its automatically keep track of building enabled it to experiment with
approach, Microsoft has emphasized occupancy and other key parameters. three different analytics engines
six design principles to overcome Weather and utility information is and determine what features were
the most common barriers, while gathered from third party providers. most important for its needs.
maximizing the program’s impact. This collection of contextual
These principles underpin the smart information is necessary for Present results in a consumable
building solution and its objectives. normalizing energy consumption data, and actionable form
allowing for a better understanding
Identify, collect and aggregate Smart building solutions can produce
of current performance and future very large volumes of complex data.
relevant data potential. It also allows for demand To interpret the data, it is important
The most important data in smart forecasting and management. to have an intuitive and customizable
building architecture is collected either user interface that includes visual
from existing building management
Employ industry-leading
features, such as editable charts with
systems or directly from installed analytics to identify savings drill-down functionality. Getting this
control panels and energy meters. A smart building solution’s value right can make a major difference in
But to put this data into context, a depends directly on its analytics enabling engineers to quickly identify
range of additional private and public engine. Algorithms that detect irregularities. For example, by using
data is also needed. This includes faults, prioritize alarms and identify Microsoft’s Silverlight technology
building layouts, occupancy levels, optimization opportunities amid vast or the new HTML5 standard, rich

16

browser-based user experiences can Keeping engineers engaged is essential Avoid disruptive change
be created that work across multiple to a smart building deployment,
devices. This can provide engineers especially during an introductory As an additional analytics layer rather
with user-friendly access to critical phase, when the analytics engine is than a replacement of an existing
information wherever they are, on PCs being fine-tuned. For smaller firms, system, a smart building solution
and mobile devices. a remote monitoring provider that constitutes a low-risk IT project with
communicates directly with on-site little disruption to ongoing business.
Centralize monitoring operations engineers can be an alternative. Given its cross-organizational
nature, smart building programs
A key implementation challenge is the
fact that building engineers often lack
Engage the organization need strong project management
functions and executive buy-in. New
the time to familiarize themselves with One aim of a smart building program
tools come with a learning curve
analytics tools and make use of them should be to influence occupant
requiring training and expectation
in their daily routine. For large-scale behavior, by providing employees and
management. With appropriate
deployments, one solution is to set other stakeholders with information
change management efforts, the
up a central operations center that about their energy footprint. Visual
adoption of the new toolset can be
connects to engineers via their PCs benchmarks or graphical renderings
accelerated and processes adjusted.
and mobile devices. In this center, showing consumption trends help
Running an extensive pilot program
dedicated employees monitor the make such metrics understandable
helped Microsoft’s organizations
whole real estate portfolio, finding for employees and management, and
get familiar with the technological
and prioritizing faults and dispatching drive behavior change. Microsoft
and operational aspects of smart
building engineers accordingly. This uses SharePoint to publish metrics
buildings prior to full rollout.
is less disruptive to the engineers’ internally, allowing tailored dashboards
role, and adds value by focusing their for specific user groups.
attention on high-priority issues.

17

4
Future smart building opportunities

18

4.1 A cloud-based 4.2 Automation and 4.3 Integration with
approach real-time analytics utilities and city
Cloud computing is set to transform We anticipate that the next generation infrastructure
information technology, by making of smart building solutions will allow
With increasing adoption of
third party applications readily organizations to automatically adjust
smart building solutions, the built
available as a service over the internet. building controls based on real-time
environment will achieve new
For smart building solutions, this will data. For example, by monitoring the
efficiencies in energy use and
deliver several benefits: security badge access information for
improvements in occupant comfort.
a building (as a proxy for the number
• Accessibility: Aggregating and But this is only part of the story.
of employees present), HVAC systems
analyzing data from disparate sources Electricity grids are being upgraded
could be automatically adjusted to
is at the core of any smart building with intelligent controls and two-way
account for increased or decreased
solution. The cloud is ideally suited communication. As individual nodes of
conditioning requirements. As an
to providing a universally available the smart grid, buildings will become
alternative, location and presence
platform for managing building active participants in managing energy
data from laptops or mobile phones
data mashed up with contextual demand and supply in a connected
could be used. Such solutions will
information and made accessible environment that includes power
rely on real-time analytics applied to
for a variety of users and devices. plants, transmission infrastructure
incoming data streams, along with
and even electric vehicles.
• Scalability: With thousands of complex event processing, to execute
sensors and controls, modern buildings automated adjustments to building For example, if a substantial share
generate large volumes of data. controls. Microsoft StreamInsight, of Microsoft’s employees were to
Microsoft’s main campus generates a component of the SQL Server use electric cars that are plugged
about half a billion data records per software, can accomplish this task and in during the day, the campus could
day from two million data points across is part of one pilot vendor’s solution. use the combined battery capacity
118 buildings. As the data volume to lower peak demand drawn from
Future tools might even use machine
and diversity of sources increases, a the utility at certain times during
learning to optimize algorithms
cloud-based architecture provides the the day.17 Likewise, demand response
over time, realizing even greater
scalability required to process massive technology18 can be used to shed
energy savings. For example,
volumes of data at an affordable cost. loads in buildings when electricity
statistical analysis, simulation and
Abundant computing power allows consumption peaks, saving cost for
predictive modeling can be applied
complex modeling, such as correlating utilities and building managers.
to determine how many chillers
external temperature, cloud cover,
need to be turned on, based on As buildings become increasingly
and wind conditions with building
forecast occupancy levels and outside networked, they play a crucial role
access activity to refine heating, air
weather conditions. Researchers in the development of energy-smart
conditioning and lighting patterns.
are also working on solutions that cities that unify the concepts of
• Ease of deployment: Cloud shape electricity demand curves of resource management and information
technology can provide a secure and HVAC systems by using buildings technology on a municipal level.19 A
uncomplicated connection between as a form of energy storage and corporate campus like Microsoft’s
off-site servers and on-site equipment. applying complex algorithms to can serve as a test-bed for many
With Microsoft’s Azure Connect the optimize energy consumption technologies that will shape the
IT team was able to exchange data through the course of the day.16 sustainable cities of the future.
with the vendor solutions within
minutes, avoiding the complexities
associated with setting up a VPN.15
Once connected, the cloud’s elastic
capacity enables vendors to serve new
customers and increase the number of
buildings managed without installing
physical servers.
For two of the three vendors that
are part of Microsoft’s pilot, the
cloud already plays an essential
part in enabling a secure data
exchange through Azure Connect.
One of them hosts its solution on
a public cloud, while the others are
currently hosted on the vendor’s
premises and at Microsoft.
19

5
Conclusion

Aggregated data and principles. Most importantly, passenger vehicles.21 The related
powerful analytics that add the underlying technologies are electricity cost savings amount
“intelligence” to existing now more widely available and to US$20-25 billion.22 Quite
building infrastructure have the easier to implement. By sharing simply, firms seeking to enhance
potential to transform the way its experience with the public, their bottom line need look no
in which companies manage Microsoft hopes to contribute to further than the offices they’re
energy across their real estate the evolution of the technology sitting in.
portfolio. In particular, building and encourage other companies
engineers can be empowered to implement programs
to take a more targeted, data- of their own.
driven approach to their work The potential for information
while automation improves technology to improve building
their productivity. This delivers energy efficiency is huge. The
substantial cost savings, while Global eSustainability Initiative
helping firms achieve carbon (GeSI)20, a consortium of
reduction targets with relatively leading high-tech companies,
low capital investments. estimates that smart building
Microsoft’s smart buildings technology has the potential to
pilot program shows that reduce carbon emissions in the
while various adoption barriers US by 130-190 million tons of
remain, these can be overcome CO2 – equivalent to the annual
by following a set of key design emissions of about 30 million

20

6
Appendix
About Microsoft Accenture Smart Building Solutions Microsoft Project Sponsors
(as part of Accenture Sustainability and Contributors
Founded in 1975, Microsoft is the Services) enables commercial building
worldwide leader in software, services, owners to cost-effectively reduce Darrell Smith
and solutions that help people and energy usage and improve occupant Senior Operations Manager,
businesses realize their full potential. comfort by managing and analyzing Real Estate & Facilities
With 90,000 employees across data to drive operational efficiency.
its business divisions and global The Accenture Real Estate Solutions Josh Henretig
subsidiaries, the company generated practice helps organizations optimize Director, Environmental Sustainability
revenues of US$ 69.9 billion for the their real estate portfolios and Jay Pittenger
fiscal year ended June 30, 2011. Its leverage their significant investment Senior Director, Real Estate & Facilities
home page is www.microsoft.com. in space to support broader business
Rob Bernard
goals and objectives.
Microsoft’s Real Estate & Facilities Chief Environmental Strategist
organization is responsible for About Lawrence Berkeley
planning, delivery and operations of Accenture Authors and
National Laboratory
Microsoft’s worldwide real estate Key Contributors
portfolio, which comprises 33 million Berkeley Lab is a U.S. Department of
square feet (3.1 million m2) and over Andri Kofmehl
Energy (DOE) national laboratory that
600 facilities across 110 countries. Senior Manager, Strategy &
conducts a wide variety of unclassified
Sustainability
scientific research for DOE’s Office
About Accenture of Science. Located in Berkeley, Abigail Levine
California, Berkeley Lab is managed by Senior Manager, Real Estate Solutions
Accenture is a global management
the University of California, and the
consulting, technology services Gregory Falco
director is Dr. Paul Alivisatos. Berkeley
and outsourcing company, with Consultant, Smart Building Solutions
Lab has a total of 4,200 employees,
approximately 236,000 people serving Kreg Schmidt
including 1,685 scientists and 475
clients in more than 120 countries. Senior Executive, Smart Building
postdoctoral fellows. Its budget for
Combining unparalleled experience, Solutions
fiscal year 2011 is US$853 million.
comprehensive capabilities across all
industries and business functions, Berkeley Lab's Environmental Energy Lawrence Berkeley National
and extensive research on the world’s Technologies Division performs Laboratory Contributors
most successful companies, Accenture research and development leading to
collaborates with clients to help better energy technologies that reduce Jessica Granderson
them become high-performance adverse energy-related environmental Research Scientist
businesses and governments. The impacts. Researchers in the Division’s Mary Ann Piette
company generated net revenues Building Technologies Department Staff Scientist and Deputy of the
of US$25.5 billion for the fiscal work closely with industry to develop Building Technologies Department
year ended Aug. 31, 2011. Its home efficient technologies for buildings
page is www.accenture.com. that increase energy efficiency, and
improve the comfort, health and safety
of building occupants.

22

References 13 One of the three pilot software solutions
is hosted on Microsoft’s premises, whereas
1 Business case numbers for the deployment the other two are hosted off-site. The
across all buildings of Microsoft’s main on-premise solution does not use Azure
campus, based on actual data from the Connect as data is not transmitted
pilot phase. The cost of deployment outside Microsoft.
correlates with the number of buildings, so 14 Including on-site vendor resources.
for smaller deployments, the percentage
may stay in a similar range. For the payback 15 Virtual Private Network.
period, the time required for installation,
16 Research rock star: Efficiency star,
testing and tuning has been factored in,
Coloradobiz, 2011.
otherwise it could be less than one year.
17 Microsoft’s corporate headquarters already
2 Energy Efficiency in Buildings, World
has a large fleet of company-operated
Business Council for Sustainable
hybrid cars to transport employees between
Development, August 2009.
buildings on campus. The company recently
3 ENERGY STAR and other climate protection started installing charging stations for
partnerships, 2009 annual report, US employee-owned electric vehicles.
Environmental Protection Agency,
18 Berkeley Lab Demand Response Research
December 2010.
Center, www.drrc.lbl.gov.
4 Commercial Buildings Energy Consumption
19 The central role of cloud computing
Survey, Energy Information Administration,
in making cities energy-smart,
December 2006.
Microsoft, 2011.
5 Building energy information systems: user
20 GeSI counts more than 30 leading IT
case studies, Springerlink, Granderson, et
and communications companies among
al, May 2010.
its members, including Microsoft. More
6 Monitoring-Based Commissioning: information available at www.gesi.org.
Benchmarking Analysis of 24 UC/CSU/IOU
21 SMART 2020: Enabling the low carbon
Projects, Mills and Mathew, June 2009.
economy in the information age, United
7 Accenture Smart Building Solutions (ASBS) States Report Addendum, GeSI, 2008.
is part of Accenture’s Sustainability
22 Greenhouse Gas Equivalencies Calculator,
Services group.
US Environmental Protection Agency.
8 A Holistic Approach to Energy Efficiency in
Datacenters, Microsoft, 2010.
9 New workplace models constitute another
strategy Microsoft employs to optimize its
real estate footprint.
10 As of 2011, Microsoft’s Sustainability
Champions program included over 430
employees working in 73 different buildings
across the main campus. Its objective is
to reduce the company’s environmental
impact through employee engagement.
Newsletters, events, training, “green
bag” lunches and monthly building
energy consumption reports by floor
or wing are the key elements aimed at
reducing employee-controlled electricity
consumption.
11 For example, engineers collaborated
with Siemens to convert data from the
proprietary Apogee format.
12 Additional data elements include square
footage, floors, types of rooms, location,
and whether a building is owned or leased.

23

Microsoft Smart Buildings White Paper

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Microsoft Smart Buildings White Paper (20)

Recently uploaded (20)

Microsoft Smart Buildings White Paper