This document discusses diagnostics and fault detection methods for smart buildings. It begins by explaining the importance of diagnostics for quickly detecting and identifying faults in engineering systems. It then describes three main classes of fault detection methods for buildings: model-driven methods based on physical models, data-driven methods using machine learning on historical data, and hybrid methods. Several examples of automated fault detection methods are provided, such as for lighting systems and air conditioners. Current issues and the state of diagnostics in buildings are also summarized. In particular, it notes that while commercial products exist, widespread adoption has been limited and improvements are still needed in areas like sensor technology and handling multiple simultaneous faults.

1. ECVET Training for Operatorsof IoT-enabledSmart Buildings (VET4SBO)
2018-1-RS01-KA202-000411
Level: 2 (two)
Module: 5 Diagnostics for smart buildings
Unit 5.1 The role and importance of diagnostics
in buildings

2. L2-M5-U5.1 The role and importance of
diagnostics in buildings
• UNIT CONTENTS
– Faults and vulnerability of engineering systemsto different
kinds of faults.
– Importance of possibilities to detect faults quickly, and to
identify their causes, severity, and consequences.
– Fault detection and diagnostics methods and their broad
classes in engineering systems.
– Fault detection and diagnostics in intelligent buildings: role
and importance
– Main classes of fault detection and diagnostics methods for
buildings and intelligent buildings.
https://pixabay.com/illustrations/business-
search-seo-engine-2082639/

3. Facility management in smart buildings
• The International Facility Management Association
(IFMA) states that facility management includes
principles of business administration, architecture,
humanities and technical sciences.
• Its aim is to achieve the long-term sustainability of
the use of the building – the optimal arrangement
and use of the interior space, ensuring the
functional and efficient functioning of the technical
equipment, the quality of the internal environment
and other user needs.
https://pixabay.com/photos/smart-home-
computer-internet-canvas-3148026/

4. Owners vs. Users of Smart Buildings
• The needs of owners and users of smart buildings partly
conform and are partly contradictory [1].
• Owners’ preferences particularly include:
– minimizing acquisition costs;
– minimizing operating costs;
– minimizing energy costs;
– minimizing repair and reconstruction costs;
– maximizing the return on investment of the building;
– continuous conservation or increase in construction value
(investment protection);
– maximizing the level of security of the building and its users.
» [1] J. Pašek, V. Sojková, Facility Management of Smart Buildings, Int. Rev.
Appl. Sci. Eng. 9 (2018) 2, 181–187, DOI: 10.1556/1848.2018.9.2.15
https://pixabay.com/photos/smart-home-
computer-internet-canvas-3148026/

5. Owners vs. Users of Smart Buildings
• The user (tenant) of the building in their own
interest does not need to interfere with the
previously-mentioned interests of the owner,
but he also prefers:
– flexibility of leased spaces;
– the quality of the indoor environment that
contributes to achieving the desired comfort of use
or to increasing work productivity;
– maximizing security level.
https://pixabay.com/photos/smart-home-
computer-internet-canvas-3148026/

6. Faults in engineering systems
• Engineering systems ranging from a simple component to a whole
complex system are vulnerable to different kinds of faults [2].
• Faults may cause sub-optimal operation and decline in
performance if not even preventing the whole system from
functioning.
• It is therefore important to detect faults quickly, and to identify
their causes, severity, and consequences.
» [2] Lazarova-Molnar S. et al, Fault Detection and Diagnosis for Smart Buildings: State of
the Art, Trends and Challenges, 2016 3rd MEC International Conference on Big Data and
Smart City.

7. The role and importance of diagnostics
in buildings
• FDD methods are proposed to address these
issues.
• FDD methods can be divided into the following
five broad categories:
– Model-Based FDD methods,
– Signal-Based FDD methods,
– Knowledge-Based FDD methods,
– Active FDD methods and
– Hybrid FDD methods.
https://pixabay.com/photos/problem-
technical-issues-technology-1951987/

8. The role and importance of diagnostics
in buildings
• FDD methods can be divided into the following five broad
categories:
– Model-Based FDD methods,
• Model-based FDD were proposed decades ago.
• Suitable modeling is a prerequisite for a successfulFDD by Model-based FDD
methods.
• Suitable models for FDD are usually obtained by first principles modeling or
systems identification methods.

9. The role and importance of diagnostics
in buildings
• FDD methods can be divided into the following five broad
categories:
– Signal-Based FDD methods,
• Signal-based FDD methods mainly use signals, which are obtained from
measurementsfor diagnostics.
• The algorithms within this category derive symptoms of a healthy system as
an output of the symptom analysis and the knowledge of the system, which
are at disposal.
• When a system is faulty, symptoms that appear in the measured signal differ
from those of healthy systems.

10. The role and importance of diagnostics
in buildings
• FDD methods can be divided into the following five broad
categories:
– Knowledge-Based FDD methods,
• Knowledge-Based FDD methods require sufficient amount of historic data.
• These methods use methods from artificial intelligence to extract the
knowledge based from the historic data reflecting the relationship between
system variables.
• The behaviour of the system is monitored in real-time and is compared with
the knowledge base to detect possible deviations and make fault diagnosis
decisions.

11. The role and importance of diagnostics
in buildings
• FDD methods can be divided into the following five broad
categories:
– Active FDD methods
• Active FDD methods are methods that evaluate the behaviour under suitable
input test signals for FDD.
• Injecting test signals in active FDD increases fault detectability, which is an
important advantage of active FDD.
• The added input test signals may compromise the performance and therefore
their effect should be minimized.

12. The role and importance of diagnostics
in buildings
• FDD methods can be divided into the following five broad
categories:
– Hybrid FDD methods.
• They are combinations of the previous.

13. Faults in smart buildings
• One of several thorough investigations [3] concluded that
typical faults in commercial buildings consist of 13 types of
faults.
• Further investigation shows that faults due to the wrong
configuration are also a typical fault in new buildings, which
has not received an adequate level of attention.
• Examples of such faults are:
– Wrongly configured building equipment, where the setting of
the equipment is wrong.
– Misplaced or wrongly wired sensors and actuators.
» [3] K. W. Roth, D. Westphalen, M. Y. Feng, P. Llana, and L. Quartararo, "Energy impact of commercial building
controls and performance diagnostics: market characterization, energy impact of building faults and energy
savings potential," Prepared by TAIX LLC for the US Department of Energy. November. 412pp, 2005.
https://pixabay.com/photos/ethics-
right-wrong-ethical-moral-2991600/

14. The annual impact of faults in buildings in terms of
energy consumption
(from Lazarova-Molnar S et al,FaultDetection and Diagnosisfor SmartBuildings:Stateof the Art, Trends and Challenges,
2016 3rd MEC International Conference on BigData and Smart City.)

15. Smart buildings FDD
• Fault Discovery and Diagnostics (FDD) is an important matter in
smart buildings.
• Operation cost of buildings has been shown to be significantly
reduced by utilizing automated FDD [4].
• These statistics have triggered a significant amount of research in
the field of FDD for buildings.
» [4] Sinopoli J., Advanced Technology for Smart Buildings, Artech House, 2016.

16. Smart buildings FDD
• While FDD can be used for other building systems, it often focuses
primarily on HVAC systems [4].
• HVAC systems are one of the more complex and energyconsuming
systems in a building involving different processes and the
interaction of different types of equipment.
• The performance of an HVAC system is measured in several
different ways:
– indoor air quality, energy consumption, and thermal comfort.
» [4] Sinopoli J., Advanced Technology for Smart Buildings, Artech House, 2016.

17. Fault detection and diagnostics in smart buildings
• Generalized at a high level, FDD methods for buildings can be
classified in three main classes:
– Model-driven methods, that are for buildings developed solely on
basis of physical models, where relations are strictly quantitatively
described.

18. Fault detection and diagnostics in smart buildings
• Generalized at a high level, we can conclude that FDD methods
for buildings can be classified in three main classes:
– Data-driven methods, thats derive the relationships and predictive
models based on historical and ongoing data collection (typically,
machine learning algorithms)

19. Fault detection and diagnostics in smart buildings
• Generalized at a high level, we can conclude that FDD
methods for buildings can be classified in three main
classes:
– Hybrid methods, that feature methods that represent a
combination of model- and data-driven approaches. combine
elements from both data-driven and model-driven methods.

20. Some Issues and Concerns in Implementing FDD in
Smart Buildings
• Some FDD Issues in Smart Buildings [4]:
– Lack of Data:
• FDD needs data from the BAS systems. If there are not enough sensors, the sensors
are inaccurate, or the building has a legacy control system, there can be issues with
obtaining the data required.
– How to Handle the FDD Information:
• Facility Management organizations need to decide how best to handle the FDD
information. A fault indicates that the system may be operational, but, is not
performing optimally.
» [4] Sinopoli J., Advanced Technology for Smart Buildings, Artech House, 2016.

21. Some Issues and Concerns in Implementing FDD in
Smart Buildings
• Some FDD Issues in Smart Buildings [4]:
…
– Alternative Ways to Deploy FDD:
• At some point in the future control manufacturers will integrate FDD routines into
their controllers, starting with the large equipment such as chillers.
– Lack of Applications For Emerging Systems:
• FDD routines do not currently address newer on-site energy sources such as solar,
wind or geothermal, or touch on power management or demand response.
» [4] Sinopoli J., Advanced Technology for Smart Buildings, Artech House, 2016.

22. Assessment of Faults in Smart Buildings
• Fault evaluation (or impact assessment) is
one of the major steps in Automated Faults
Detection and Diagnosis (AFDD) process.
• The severity of the fault and its impact on
energy consumptionis essential for
prioritizing the repair.
https://pixabay.com/illustrations/dete
ctive-searching-man-search-1424831/

23. Assessment of Faults in Smart Buildings
• However, assessing the impact (energy and
cost) or the severity of the fault is difficult
because in many cases the information needed
to make the assessment is not easily available.
• The fault impact can be used to prioritize the
repairs, which will result in reduced energy and
costs, improved comfort and equipment life,
and reduced service costs.
https://pixabay.com/illustrations/dete
ctive-searching-man-search-1424831/

24. Examples of automated FDD in Smart Buildings
• Automated FDD method can be used for
monitoring the faults of lightingsystems [5].
• The AFDD method uses the light intensity sensor
to measure the approximatelevel of the light and
its statistical distribution.
» [5] Woohyun Kim & Srinivas Katipamula (2017):A Review of
Fault Detection and Diagnostics Methods for Building Systems,
Science and Technology for the Built Environment, DOI:
10.1080/23744731.2017.1318008
https://pixabay.com/photos/light-bulb-
idea-creativity-socket-3104355/

25. Examples of automated FDD in Smart Buildings
• The detection thresholds are identified at the point
where the output of light intensity sensors starts to
decrease.
• Furthermore, real-time health monitoring is
conducted based on input current, input voltage, and
board temperature to predict the light output power
degradation of an LED in real-time [5].
» [5] Woohyun Kim & Srinivas Katipamula (2017): A Review of Fault
Detection and Diagnostics Methods for Building Systems, Science
and Technology for the Built Environment, DOI:
10.1080/23744731.2017.1318008
https://pixabay.com/photos/light-bulb-
idea-creativity-socket-3104355/

26. Examples of automated FDD in Smart Buildings
• AFDD method is also developed for air-
conditioners and heat pump systems that have a
fixed-speed and a variable-speed compressor [5].
• Under this method, five different faults are
detected: 1) loss of compressor performance, 2)
low or high refrigerant charge, 3) fouled condenser
or evaporator filter, 4) faulty expansion device, and
5) liquid-line restriction.
» [5] Woohyun Kim & Srinivas Katipamula (2017):A Review of Fault
Detection and Diagnostics Methods for Building Systems, Science
and Technology for the Built Environment, DOI:
10.1080/23744731.2017.1318008.
https://pixabay.com/illustrations/air-
conditioning-air-conditioner-3679756/

27. Examples of automated FDD in Smart Buildings
• The performance models for capacity and power
consumption for normal conditions are developed
to estimate the expected reference value.
• A comparison between current estimated
performance and normal expected values is used to
determine whether a fault, when detected, is
severe enough to justify service [5].
» [5] Woohyun Kim & Srinivas Katipamula (2017):A Review of Fault
Detection and Diagnostics Methods for Building Systems, Science
and Technology for the Built Environment, DOI:
10.1080/23744731.2017.1318008.
https://pixabay.com/illustrations/air-
conditioning-air-conditioner-3679756/

28. Current state of diagnostics in buildings
• There are more commercial AFDD products and services
available in the market.
• However, the penetration is not as widespread as one would
expect [5].
• Low-cost reliable sensing for some type of measurements
(air flow, pressure,power, etc.) are still lacking.
• Improvement in the last decade is the development of low-
cost AFDD algorithms that reduce the number of sensors
necessary to detect a set of faults or degradation of the
performance of a system.
» [5] Woohyun Kim & Srinivas Katipamula (2017):A Review of Fault Detection
and DiagnosticsMethods for Building Systems,Science and Technology for
the Built Environment, DOI: 10.1080/23744731.2017.1318008.
https://pixabay.com/photos/pros-and-
cons-weigh-compare-baskets-
2028471/

29. Current state of diagnostics in buildings
• Some recent studies have also been able to detect multiple
simultaneous faults, while earlier AFDD methods did not
handle multiple faults that occur simultaneously.
• Occasionally, the faulty component causes faults in other
system components and the AFDD method must be able to
diagnose all fault sources simultaneously.
• If only one fault is diagnosed and repaired, the system will
continue to operate with an undiagnosed fault that could
cause the repaired component(s) to fail again.
https://pixabay.com/photos/pros-and-
cons-weigh-compare-baskets-
2028471/

30. Current state of diagnostics in buildings
• Some recent AFDD methods pfocused on the
selection of a suitable threshold to prevent
frequent fault alarms.
• High false alarm rates and a lack of good
threshold selection strategies prevent building
industry from embracing the latest AFDD
strategies.
https://pixabay.com/photos/pros-and-
cons-weigh-compare-baskets-
2028471/

31. Current state of diagnostics in buildings
• If thresholds were set too close to normal conditions,
the AFDD system would be too sensitive, which
would lead to false alarms.
• If thresholds were set too far from normal
conditions, the AFDD system would miss faults that
potentially could reduce system performance.
• Therefore, it is important to define reasonable
thresholds so that the presence of fault is detected.
https://pixabay.com/photos/pros-and-
cons-weigh-compare-baskets-
2028471/

32. Thank you for your attention.
https://pixabay.com/illustrations/thank-you-polaroid-letters-2490552/

33. Disclaimer
For further information, relatedto the VET4SBO project, please visit the project’swebsite at https://smart-building-
operator.euor visit us at https://www.facebook.com/Vet4sbo.
Downloadour mobile app at https://play.google.com/store/apps/details?id=com.vet4sbo.mobile.
This project (2018-1-RS01-KA202-000411) has been funded with support from the European Commission (Erasmus+
Programme). Thispublicationreflects the views only of the author, and the Commission cannot be held responsible
for any use which may be made of the informationcontainedtherein.