COMPLEX EVENT PROCESSING IN BUILDING
MANAGEMENT SYSTEMS — A PROTOTYPE
System architecture overview Building and technology
passport
Passports are spatial databases containing
information about facilities and devices
that university uses. Each item in
the databases is identified with unique
code with hierarchic structure.
Building management
data semantics
Joining data from BMS and from passport
allows as to add „machine-readable“ information
about semantics (meaning) of
data points (i.e. inputs, outputs, control
variables) in BMS. In addition to the information
from technology passport we add
information about measured quantity
(temperature, pressure, humidity, actual
current, voltage...) of each data point.
Adam Kučera, Faculty of Informatics, Masaryk University
Introduction
Complex Event Processing (CEP) is set of tools and techniques allowing us to handle events from large systems in real time. CEP in commonly used in network security, algorithmic stock-trading, fraud
detection or business process monitoring for correlating, aggregating and analyzing various events.
Modern ("intelligent") buildings are often equipped with BMS – network of interconnected devices that ensures integrated building operation, control and monitoring. Building operators and maintenance
staff are easily overwhelmed by large amounts of data and events produced by the BMS. In this way BMS are similar to other fields where CEP is successfully used (network monitoring, credit
card fraud detection, algorithmic stock trading etc.).
Test environment — BMS at Masaryk University
 Integrates monitoring and control of University Campus in Brno—Bohunice (first phase of
construction ended in 2007) and other university buildings in Brno.
 Uses BACnet as common communication protocol
 Over 700 native BACnet devices by approximately 10 vendors
 Integrated systems/technologies: HVAC (Heating, Ventilation and Air-conditioning), Fire
alarms, Security system, Access control system, CCTV, Power systems, UPS monitoring, university
information system (controlling door locks according to lecture room timetables)
The goals
Aim of this research is to adapt existing CEP framework (Microsoft StreamInsight) to
work with data from intelligent buildings and provide building operators with the easily
understandable outputs that help to optimize building operation and detect various
types of faults that occurs in a building. The CEP framework was extended to be able to
work with specific features of building management data (location, source device,
measured/controlled quantity). A prototype of the system was implemented and tested
in real operation environment.
Attribute Value Meaning
Data point 102.AV145 ID of data point
in BMS
Location BHA12P01005 Room 005 in the
basement of
building 12 at
university camDevice
BAPK Energy meter
Quantity ESS0 Electricity con-
sumption
Selecting, filtering, joining,
aggregating data
Using building management data semantics,
we are able to perform selections, filtering
and computations (average, summa,
maximum, minimum) over the data from
the BMS. The system uses time windows for
performing continuous computations of the
data characteristics. We can utilize hierarchy
of passport databases and use masks to
gather related data into separate groups.
System prototype — Data analysis and visualization
System runs continuous user-defined queries that transform input data from BMS using various methods (selection, filtering, joining,
aggregation, grouping). Results of the computations performed by the system can by further analyzed automatically (e.g. Detecting
fault on particular device) or stored for later visualization that can be examined by users.
References
D. Luckham, "The Power of Events", Addison-Wesley,
2002.
H. Merz, T. Hansemann and C. Hübner, "Building automation",
Springer, 2009.
O. Etzion and P. Niblett, "Event Processing in Action",
Manning Publications, 2010.
A. Kučera, "Complex Event Processing in Building Management
Systems", Master thesis, Masaryk University,
Faculty of Informatics, 2012.
DISTRIBUTED EVENT-DRIVEN MONITORING MODEL
FOR CLOUD DATACENTERS
BotanickáDaniel Tovarňák Tomáš Pitner
C
C
C
C
1
1
2
3
P
P
P
P
Introduction
References
Research Goals
Distributed Event-driven Monitoring Model
1
WASL’08, USENIX Association,
2 Event Processing in Action
3 ňá
SYNASC’12
PR
PR
PR PR
publish subscribe
P roducer C onsumerPR ocessing agentS ensor
(Virtualized)
Hardware
Application
Operating
System
Middleware
User
1
2
1
2
3
P
behavior &
state events
complex
events
simple
subscriptions
complex
subscriptions
P
P
P
P
Conclusion and Expected Results
Distributed Event-driven
Monitoring Architecture
?
3
Methodology
PR
1|2|3 – access
permisions
Aims and Objectives
Distributed Eventdriven
Monitoring Architecture
2
1 1
Scope
Developing the
(project at FI MU)
ICT partners
network
Platform for Industrial Cooperation
Platforma průmyslové spolupráce
Project partners:
 IBM Česká republika, spol. s r. o.
 Microsoft s. r. o.
 Red Hat Czech s. r. o.
 JIC, zájmové sdružení právnických osob
This project is co-financed by the European Social Fund
and the state budget of the Czech Republic
Contact information
http://lasaris.fi.muni.cz/pps
pps@fi.muni.cz
Project lead: Tomáš Pitner
Project manager: Stanislava Sedláčková
CZ.1.07/2.4.00/17.0041
The Platform project addresses those requirements by
promoting four key activities that are interconnected
through the two-phased internship model:
 student projects – team or individual ▪ related to
curricula or student's interests ▪ projects are lead by
faculty supervisors and consulted by company experts
 workshops – orientation on technology and business
→ student projects and workshops together create a
first phase of the internship model allowing students
to meet experts from companies at the faculty
 internships at companies – correspond to curricula
and student projects ▪ with support of the faculty
 → internships are a second phase – they offer
students opportunities to use their knowledge they
have gained by participating in first-phase activities
 own business – support for start-ups and spin-offs ▪
incubation at South Moravian Innovation Centre (JIC)
with professional support from the faculty
From idea to realisation
The faculty currently cooperates with
almost 30 companies in area of ICT.
Cooperation with industry is one of
the key instruments to fulfil FI's
vision: '… to become an outstanding
research university with significant
accomplishments in this area...'.
A sustainable approach to
developing partners network has to
meet some requirements like:
a reasonable financial burden,
simplicity of implementation of such
activities and their attractiveness.
Q: Best approach?
Q: What is in it for me?
FI – Faculty of Informatics, S – student, C – company
Benefits FI S C
cooperation as a
source for new ideas   
student as an agent to
transfer knowledge   
distribution of the
“resources load”   
start-ups and spin-offs
as a future FI partner   
?
at FI
Workshops Student projects
Internships
at
companies
start-ups
&
spin-offs
Design of a system for the analysis
of social media content
Introduction
Methodology
Conclusion
References
• Boyd, D. M., Ellison, N. B. (2007). Social Network Sites: Definition, History, and Scholarship. Journal of Computer-Mediated
Communication, 13(1), p. 210-230.
• Pang, B.; Lee, L. (2008). Opinion Mining and Sentiment Analysis. Foundations and Trends® in Information Retrieval. 2(2), p. 1-135.
• Ministr, J., Ráček, J. and Toth, D. (2012) ‘Visualization of the discussion content from the Internet’, IDIMT- 2012 ICT Support for
Complex Systems – 20th Interdisciplinary Information Management Talks, Jindřichův Hradec, Linz: Trauner Verlag, p. 297-304.
Jaroslav Ráček
Jaroslav.Racek@ibacz.eu
Faculty of Informatics, Masaryk University & IBA CZ
Brno, Czech Republic
Active users of social media produce large volumes
of data on a daily basis. This data could contain
patterns and expression of sentiments which have
value to commerce and academics alike.
Many algorithms and methods for analysing
social media data exist, but there is not any one
platform uniting these tools and services.
The solution is a design of a single system with four layers, to
integrate all required elements. These layers are the: collection
of data from different sources, data management and storage,
analytical algorithm usage focused on relevant aspects and
visualisation of analytical results and source data.
This paper introduces a design of new
component based system for social media
analysis, which can be easily configured for use in
academic and also commercial environment in
traditional areas of social media analysis. The
system was designed on the basis of applied
research in cooperation with industrial partners
of FI MU.
• Chronologically arranged posts vs. threads of
posts
• Expressions of opinions, attitudes and
reactions that can be viewed by other users.
• Indirect conversation with benefit of
hindsight
• Integration of all components in a single platform
• Provides human and software interface for visualisation of results
and access to retrieved data
• Specific configuration for different usage of the system
• Direct and indirect conversations, sharing of profile
content
• Mainly short messages, no general topic
• Long main article and related short comments
• General or focused on narrow area of interest
• Adapters for different data sources
• Different purposes: personal diary, significant
events of life, info about interests, company sites
• Interactive (commented) or static sites
• Specialized solutions for management of massive
amounts of data and running indexing on it
• Possibility of initial filtering of data
• Essential task for text analysis
• Identification of key terms contained in a source
text
• “Social” network of users, groups, similar
behaviour
• General part of any specific analytical tool
• Visualisation of analysis results in a form of
different types of graphs and record details
• Identification of multiple identities as the
same person
• Using behaviour patterns, timing of posts,
typos, phrases
• Evaluation of text combinations by comparison
to records in knowledge base
Dalibor Toth
xtoth2@fi.muni.cz
&
Graphics design of a mobile UI
for primary school
administration
References:
Cooper, Allan. 2007. About face 3, Wiley Publishing, Inc.
Tidwell, Jenifer. 2005. Designing Interfaces, O'Reilly Author: Lenka Plháková
Semantically Partitioned Peer to Peer
Complex Event Processing
Exploiting Information Loss
References
[1] Nguyen F., Pitner T. 2012. Information system monitoring and notifications using complex event processing.
In Proceedings of the Fifth Balkan Conference in Informatics (BCI '12). ACM, New York, NY, USA, 211-216.
[2] Kunc P., Nguyen F., Pitner T. 2013. Towards Effective Social Network System Implementation.
New Trends in Databases and Information Systems Advances in Intelligent Systems and Computing. Springer Berlin Heidelberg, 327-336.
[3] Nguyen F., Škrabálek J. 2011 NotX service oriented multi-platform notification system.
In Computer Science and Information Systems (FedCSIS). Szczecin, Poland, 313-316.
[4] Wu, E., Diao, Y., Rizvi, S. 2006. High-performance complex event processing over streams.
In Proceedings of the 2006 ACM SIGMOD international conference on Management of data - SIGMOD ’06.
[5] Akram, S., Marazakis, M., Bilas, A. 2012. Understanding and improving the cost of scaling distributed event processing.
In Proceedings of the 6th ACM International Conference on Distributed Event-Based Systems (DEBS '12). ACM, New York, NY, USA, 290-301.
[6] Luckham, D. C., Frasca, B. 1998. Complex Event Processing in Distributed Systems.
In Standford University, Vol 28.
Typical Complex Event Processing:
Red produceres are sending events
to black Complex Event
Processing engine.
Scaling Complex Event Processing [6] (CEP) applications is
inherently problematic. Our solution for scaling CEP applications
is fully distributed and aspires to scale CEP to the limits of
current hardware. Our solution simplifies existent Event
Processing Network abstraction and adds features on the level of
CEP that change direction of its usage.
Complex Event Processing was introduced by David Luckham. We are mainly
concerned with subarea of Luckam's work related to distributed CEP [6] (also
studied by [6] and [4]).
Motivation of our work stems from our work related to event processing [3]. We
have applied our theoretical ideas in concepts introduced in [2] and gave brief
introduction to our overall research in [1].
Results
We believe that fully distributed peer to peer CEP is
inevitable solution to high volume event streams. Our
implementation of presented concept is called peer CEP
(PCEP). The main property of PCEP is semantic scaling.
The scaling is not done by brute force or by exploiting
specific feature of specific event context, but it is done by
exploiting partitioning of peers according to their's affiliation
to matching rules.
The developed distributed engine is written in Java and thus
runs on heterogenous platforms. In the implementation we
leverage distributed algorithms developed in theirs natural
form - not optimized to the state of being obfuscated code.
In theoretical point of view, our solution introduces
rigorously defined trade off between matching capabilities
and throughoutput of the events. In the future we plan to
extend this knowledge by revealing statistical properties of
mentioned trade off situation.
Masaryk University
Faculty of Informatics
Botanická 68a
602 00 Brno
Czech Republic
Related WorkAbstract
There is ongoing research to distribute CEP. Every author
makes his own definition of distributed CEP. Usually, it
refers to a use of filters on producers or parallelizing
existing CEP operators. We see distributed CEP
differently. We aim to distribute the processing at
semantic level. We do not want to just filter unknown
events. We allow users to leverage standard operators
and give them framework to easily trade off processing
power with matching precision.
Distributed CEP
P1
P2
P3
P4
P5
EventEvents are
traveling on
edges
towards
engne.
Event
The definition of an event varies
based on context of CEP. However
some parameters are always the
same. Each event has defined time
of creation and producer.
Events should be as fine grained as
possible - to allow effective CEP.
That means thousands, even milions
of events per second are desirable.
This is not uncommon thing today
with advent of social networks, faster
networking hardware and computer
driven high frequency trading.
Filip Nguyen
xnguyen@fi.muni.cz
P1
P2 P4
P5
Suppose we know that P1 and P3
produce events at the same time with
high probability
Then we can add an engine between
them and match events.
Very simple query that matches
events that happenend in the time
window of 0.01 second.
select name(EA), name(EB)
where abs(time(EA) - time(EB))
< 0.01s
&& EA!=EB
This query needs all the produced
events.
P1
P2 P4
P5
12.24
13.11
14.44
12.27
12.29
14.00
16.24
12.20
22.24
12.24
13.11
14.44
12.29
14.00
16.24
12.20
22.24
12.27
Here the second engine between P1 and
P3 will match events and is loaded with
less events than the former engine.
Unfortunatelly the event produced at
12.20 by P5 will not be matched. This is
the trade off situation in our solution.
How to deploy the engines dynamically?
Our solution is to turn each producer into
an engine. This way we gain additional
property - high availability.
We refer to this model as peer to peer model.
The events are distributed throughout the
formed network. Some of the events travel on
dedicated paths, some are broadcasted. This
behavior is based on the result of partitioning
algorithm.
A node is said to be a peer.
There is another result we present - partitioning algorithms.
We believe those algorithms may be extended and
generalized to be used in other fields for set partitioning and
analysis of data sets. These algorithms join several
Distributed algorithms, Statistics and Complex Event
Processing. We theorize, that the partitioning may be done
in a distributed fashion.
We also believe in adoption by users. We strive robust
architecture. Our solution is Open Source and we plan to
apply for Apache Incubation. We believe the science should
be done for greater good and sharing the code will improve
the implementation.
Lastly, our solution is not mutually exclusive with recent
research in the area of CEP. It will be possible to use
standard CEP engines on the peer nodes and thus
augmenting existing tools with PCEP.
?
Event
Coarse
Event
Peer Network
Partitioning Algorithm
Coarse Grained
Event
CEP Based and
Monte Carlo Algorithm
Basic Approach
Links
Github: https://github.com/nguyenfilip/pcep
LaSArIS: http://lasaris.fi.muni.cz/
LinkedIn: http://www.linkedin.com/pub/filip-nguyen/27/60/5b4
University: http://www.fi.muni.cz
P3
P3
Acknowledgments
Thanks to all the staff members of the Masaryk University Department of Building Passportization for providing
maps and consultations. Also, great thanks to Michal Holčík and Adam Berthóty for implementing localization
prototypes.
References
[1] A. Chen et al: “An algorithm for fast, model-free tracking
indoors”, SIGMOBILE MCCR, vol. 11 no. 3, 2006
[2] S. Cho: “MEMS Based Pedestrian Navigation System”, JN,
vol. 59, pp. 135–153, Royal Institute of Navigation, 2006
[3] M. Arulampalam et al: “A Tutorial on Particle Filters for
Online Nonlinear/Non-Gaussian Bayesian Tracking”, SP,Online Nonlinear/Non-Gaussian Bayesian Tracking”, SP,
vol. 50 no. 2, 2002
Preliminary Results
• implemented prototype Android application
• results accurate to 2.3 meters
Dead Reckoning
• calculating position using previously determined
coordinates advanced by known speed and course
• position tracking - calculating steps
• step length estimated using neural network [2]
• course determined by gyrocompass.
Sequential Monte Carlo FilteringSequential Monte Carlo Filtering
• particles evenly spread in probable location
determined by RSS fingerprint database
• particles set to motion by step detection events
• eliminated particles, which hypothetical motion leads
through impassable obstacles
• results in improvement of location estimation.
Methodology
• localization system based on [1, 2, 3]
• consists of three localization techniques
• techniques merged together for more accurate results
Wi-Fi Localization
• Received Signal Strength fingerprinting
•• used to create a database of APs and their RSS
• mapped to location coordinates
• receiver’s location estimated from SS maps
Introduction
• difficult navigation in complex buildings
• even if equipped with signs
• no direct visibility to GPS satellites
• GPS tracking not possible
Jonáš Ševčík
Indoor User Localization
Using Mobile Devices
Implementation of the
(project at FI MU)
internship model
in ICT area
Platform for Industrial Cooperation
Platforma průmyslové spolupráce
Project partners:
 IBM Česká republika, spol. s r. o.
 Microsoft s. r. o.
 Red Hat Czech s. r. o.
 JIC, zájmové sdružení právnických osob
This project is co-financed by the European Social Fund
and the state budget of the Czech Republic
Contact information
http://lasaris.fi.muni.cz/pps/internship
pps@fi.muni.cz
Project lead: Tomáš Pitner
Internship manager: Jana Bartáková
CZ.1.07/2.4.00/17.0041
First phase

students can work on real projects or theses in
application area with support of teachers from FI
 students advance knowledge by attending
technical workshops, meetings with experts
from business and/or standard curricula lectures
 students have opportunity to meet specialists from
companies directly at FI and become familiar with
practical applications of theoretical approaches
and modern technologies used by companies
● participating in activities of the first phase may help
students to determine their own career vision
Second phase
 students go to pre-selected companies where they
can use their experience and knowledge gained
from first phase of the internship model, and
 their vision can guide them during the process of
setting strategy and goals for their internship →
thus they are able to plan and manage it
Two-phase model
Q: What is in it for me?
Whilst
company resources are
significantly limited, the effect of
internship (transfer of knowledge
and the overall contribution for all
parties) needs to be maximized
→ internship is not suitable to draw
up like “vacation job” but better
approach is to plan and
managed as a project
→ goals, time limitations, resources
has to be set → student's
vision is the key factor
Best concept?
Benefits FI S C
shaping own future ✓ ✓
establishing contacts ✓ ✓ ✓
praxis related to studies ✓ ✓ ✓
enrichment of theoretical
knowlege by experience
✓ ✓ ✓
systematic buildig and
extending knowledge
✓ ✓ ✓
source of potential
employees and employeers
✓ ✓
financial support of PPS ✓ ✓
FI – Faculty of Informatics, S – student, C - company
team projects
thesis (preliminary analysis)
+
workshops/seminars/lectures
at FI
team projects
thesis (preliminary analysis)
+
workshops/seminars/lectures
at FI
first phase
internships
at companies
internships
at companies
second phase
involve
experts from
companies
involve FI
(administrative
and technical
support)
full – time jobfull – time job
part – time jobpart – time job
Poster Presentations
Supporting the Process of Learning
Mobile Application User Interfaces
References
Lucia Tokárová, Faculty of Informatics, Masaryk University, Brno, Czech Republic
The objective of this research project is to investigate how people learn to
use mobile applications, and how can this process be supported in different
phases so that they quickly perceive the value of the application, accomplish
basic tasks, and gradually learn new features in a natural way.
•	 focus on understanding the activity and
attaining momentary goals
•	 frequent, perceptually salient errors with
immediate consequences
•	 help of parent or teacher
The process of learning
•	 gaining experience
•	 less frequent and serious mistakes
•	 level of concentration is reduced
•	 attaining an acceptable level of
performance
•	 performance is not actively controlled
•	 most people do not perceive an urge for
further improvements
•	 the same level of performance is
maintained for months/years
•	 continuous learning
•	 individuals are not satisfied with
acceptable level of performance
•	 deliberate practice: seeking challenges to
achieve ever higher level of performance
•	 first contact with application
•	 focus on understanding purpose of the
application, accomplishing basic tasks
•	 ad-hoc feature exploration
•	 asking for help (experienced user/forum)
•	 practice and familiarization
•	 short sessions with the application,
predictable usage patterns [2]
•	 ad-hoc problem solving
•	 fewer mistakes, faster task completion
•	 performance becomes autonomous,
users focus on task instead of UI
•	 most users stop learning new strategies
and start actively avoiding frustrating and
unfamiliar situations
•	 finding more efficient ways
•	 learning shortcuts
•	 exploration of advanced features
•	 personalization of user interface
•	 systematic approach to problem-solving
•	 with	practice,	users	become	quicker	but	
not	more	efficient
•	 problem-solving	strategies	differ	
among	user	groups	[2]
•	 users	do	not	perceive	an	urge	for	
further	improvements,	they	tend	to	
stick	to	familiar	strategies
•	 as	individuals	improve	their	
performance,	their	needs	evolve	and	
the	user	interface	should	reflect	these	
changes
The	process	of	learning	mobile	applications
Gestural UIs are engaging and
intuitive for simple tasks but not
for advanced operations [4].
Sessions with mobile applications
are short [1, 6] and variable in the
context of use [6], which affect
users’attention.
Solely visual user interfaces
without haptic feedback prevent
activation of the muscle memory.
Small screens provide space for
displaying high-priority functions
and reduce discoverability of
advanced features [4].
With the recent expansion of the mobile industry, applications for mobile devices are becoming more complex,
empowering people to perform more advanced tasks. However, modern mobile user interfaces introduce several
challenges, which affect learnability of mobile applications. For example:
Objective Methodology & expected results
•	 most	users	leave	mobile	applications	in	
the	initial	phase	[3]	due	to	insufficient	
onboarding	strategies
Research	question:	How to continuously support the process of learning mobile application user interfaces?
Learning is a long-term process. Individual’s needs are changing over time. To achieve the highest levels of expertise,
learners should be engaged in deliberate practice to continuously improve their performance. (via [5])
Thisstudywillconsistofobservationofusers’behaviorintheprocessoflearning
mobile UIs. Patterns in users’ behavior will be investigated, in order to create
learning profiles of representative user groups. Further profile examination
should lead to the design of support mechanisms that will encourage various
types of learners in the process of continuous learning of mobile applications.
[1] Böhmer, M. et al. 2011. Falling asleep with angry birds, facebook and kindle: a large scale study on mobile application usage. Proceedings of the 13th International Conference on Human Computer
Interaction with Mobile Devices and Services (MobileHCI 2011). Sweden, Stockholm, 30 August – 2 September 2011:47–56.
[2] Burnett, M.M. et al. 2011. Gender pluralism in problem-solving software. Interacting with Computers, 23(5):450–460.
[3] Min, A. 2011. 26% of Mobile Application Users are Fickle — or Loyal. http://www.localytics.com/blog/2011/26percent-of-mobile-app-users-are-either-fickle-or-loyal/ [05 December 2012].
[4] Norman, D. 2010. Natural user interfaces are not natural. interactions, 17(3):6–10.
[5] Oulasvirta, A. et al. 2011. What does it mean to be good at using a mobile device? An investigation of three levels of experience and skill. International Journal of Human-Computer Studies, 69(3):155–169.
[6] Oulasvirta, A. et al. 2005. Interaction in 4-Second Bursts : The Fragmented Nature of Attentional Resources in Mobile HCI. Proceedings of the SIGCHI Conference on Human Factors in Computing Systems
(CHI 2005). USA, Oregon, Portland, 2 – 7 April 2005:910–928.
The	initial	phase
PrObLeM 1 PrObLeM 2 PrObLeM 3 PrObLeM 4
The	expert	approach
The	last	phaseThe	middle	phase
The	initial	phase
Masaryk University, Brno 26 February 2013