KAUST Conference, October 5-7, 2015
HUMAN-MACHINE
NETWORKS AND
INTELLIGENT
INFRASTRUCTURE
Co-organized by Profs. Meriem Laleg and Jeff Shamma
with financial support from the KAUST Office of Sponsored Research
Co-sponsored by KAUST Industry Collaboration Program (KICP)
AGENDA
08:30 – 09:00
09:00 – 09:15
09:15 – 10:15
10:15 – 10:30
DAY 1
Monday, October 5, 2015
COFFEE & BREAKFAST (BUILDING 9 – HALL 2)
WELCOME AND OPENING REMARKS
Dr. T.A. Abrajano, Director, Office of Sponsored Research, KAUST
Keynote Lecture #1
CYBER-PHYSICAL-SOCIAL SYSTEMS AND RESILIENT INFRASTRUCTURES – A PERSPECTIVE FROM NSF
Dr. Pramod Khargonekar, Assistant Director for Directorate of Engineering US National Science Foundation (NSF)
BREAK
10:30 – 11:10
CONNECTING DISTRIBUTED CONTROL AND DISTRIBUTED OPTIMIZATION IN THE POWER GRID
AND GENERAL NETWORK SUSTEMS
Prof. Na Li, Harvard University
11:10 – 11:50
DECENTRALIZED AND OPTIMAL CONTROL OF INTER-AREA OSCILLATIONS IN POWER NETWORKS
Prof. Florian Dorfler, Swiss Federal Institute of Technology (ETH)
13:10 – 13:50
DEMAND RESPONSE: A VALUABLE ASSET FOR A SMARTER ELECTRIC GRID
Prof. Ali Alawami, King Fahd University of Petroleum and Minerals (KFUPM)
13:50 – 14:30
APPLIED ROBOTICS IN THE ENERGY SECTOR
Mr. Fadel Abdellatif, Saudi Aramco
15:30 – 16:10
LAYERED CONTROL OF ROAD FREIGHT TRANSPORT
Prof. Karl Johansson, Royal Institute of Technology (KTH)
11:50 – 13:10
13:50 – 14:30
15:10 – 15:30
16:10 – 16:50
18:30
LUNCH CAMPUS DINER (RESERVED SECTION)
TOWARDS OIL AND GAS INFRASTRUCTURE SUSTAINABILITY ONSHORE AND OFFSHORE
Prof. Mansour Karkoub, Texas A&M University at Qatar
BREAK
INTELLIGENT OPERATION OF LOW EMISSIONS POWER GENERATION
Prof. Ali Abbas, University of Sydney
CONFERENCE GALA DINNER YACHT CLUB RESTAURANT
Human-Machine Networks and Intelligent Infrastructure
3
AGENDA
08:30 – 09:00
DAY 2
Tuesday, October 6, 2015
COFFEE & BREAKFAST (BUILDING 9 – HALL 2)
09:00 – 10:00
Keynote Lecture #2
10:15 – 10:55
THE INTERNET IS NOT AN INTELLIGENT INFRASTRUCTURE: CAUSES AND POTENTIAL SOLUTIONS
Prof. Constantine Dovrolis, Georgia Institute of Technology
11:35 – 12:15
RESILIENT MONITORING AND CONTROL OF DISTRIBUTION NETWORKS
Prof. Saurabh Amin, Massachusetts Institute of Technology
13:30 – 14:10
DESIGN AND DEPLOYMENT OF CONTROLLERS IN PASSIVELY OBSERVED NETWORKS
Prof. Donatello Materassi, University of Tennessee
10:00 – 10:15
10:55 – 11:35
12:15 – 13:30
DYNAMIC CO-EVOLVING SOCIOTECHNICAL NETWORKS
Prof. John Baras, Lockheed Martin Chair in Systems Engineering, University of Maryland
BREAK
CITY DYNAMICS: A BIG DATA ANALYTICS PLATFORM FOR CITIES
Dr. Anas Alfaris, King Abdelaziz City for Science and Technology (KACST) & Massachusetts Institute of Technology
LUNCH CAMPUS DINER (RESERVED SECTION)
14:10 – 14:50
THE SMART OCEAN: TRANSFORMATIVE INTEGRATED OCEAN OBSERVATION NETWORKS
Prof. Burt Jones, KAUST
15:10 – 15:50
MODELING, OPTIMIZATION AND CONTROL OF SUSTAINABLE DESALINATION PLANTS
Prof. Meriem Laleg, KAUST
14:50 – 15:10
15:50 - 16:30
19:00
4
BREAK
UTILITIES OF THE FUTURE: NEW BUSINESS MODELS AND REGULATORY FRAMEWORKS
Dr. Rolando Fuentes, King Abdullah Petroleum Studies and Research Center (KAPSARC)
DINNER - BBQ PURE RESTAURANT - ISLAND RECREATION CENTER
KAUST Conference, October 2015
AGENDA
08:30 – 09:00
Wednesday, October 7, 2015
COFFEE & BREAKFAST (BUILDING 9 – HALL 2)
09:00 – 10:00
Keynote Lecture #3
10:15 – 10:55
ALGORITHMS FOR AN INTELLIGENT AIR TRANSPORTATION INFRASTRUCTURE
Prof. Hamsa Balakrishnan, Massachusetts Institute of Technology
10:00 – 10:15
10:55 – 11:35
11:35 – 12:15
12:15 – 13:30
13:30 – 14:10
DAY 3
HUMAN-MACHINE INTERACTION: RATIONAL AND PREDICTABLE FROM WHAT PERSPECTIVE?
Prof. Amy Pritchett, David S. Lewis Associate Professor of Cognitive Engineering, Georgia Institute of Technology
BREAK
DRONE VISION: APPLYING COMPUTER VISION ALGORITHMS TO A SWARM OF NETWORKED
AERIAL VEHICLES FOR REMOTE SENSING AND UNDERSTANDING
Dr. Neil Smith, FalconViz
POSTER SESSION BUILDING 9 - WEST SIDE LOBBY AREA
LUNCH BUILDING 9 - WEST SIDE LOBBY AREA/CAFETERIA
ON AUTONOMIC COMPUTING
Prof. Mariette Awad, American University of Beirut
14:10 – 14:50
DISTRIBUTED LEARNING DYNAMICS: CONVERGENCE IN ROUTING GAMES AND BEYOND
Mr. Walid Krichene, University of California Berkeley
15:10 – 15:50
DECENTRALIZED Q-LEARNING FOR STOCHASTIC DYNAMIC GAMES
Prof. Gurdal Arslan, University of Hawaii
14:50 – 15:10
15:50 – 16:30
18:30
BREAK
THE DYNAMICS OF SOCIAL INFLUENCE
Dr. Bary Pradelski, Swiss Federal Institute of Technology (ETH)
DINNER FISH RESTAURANT - THUWAL VILLAGE (BY INVITATION)
Human-Machine Networks and Intelligent Infrastructure
5
KEYNOTE TALKS
CYBER-PHYSICAL-SOCIAL SYSTEMS
AND RESILIENT INFRASTRUCTURES
– A PERSPECTIVE FROM NSF
DR. PRAMOD
KHARGONEKAR
Assistant Director for
Directorate of Engineering US
National Science Foundation
(NSF)
Abstract: The main goal of this talk is to present NSF programs in advancing
research frontiers in systems where humans and cyber-physical systems intersect.
Increasingly, pressing societal grand challenges require leveraging of advances
in cyber-physical systems. However, to be successfully meet societal needs, it
is necessary to fully integrate knowledge and insights from social-behavioraleconomic sciences. This is particularly true in infrastructure systems. Another major
area of opportunity is the services sector which includes health care, education,
government, etc. NSF Engineering Directorate is enabling the research community
to address these challenges and create new foundational knowledge through
a variety of programs and initiatives. I will give an overview of major challenges,
relevant NSF programs and emerging trends and opportunities.
BIOGRAPHY
Dr. Pramod P. Khargonekar was appointed by the National
Science Foundation (NSF) to serve as Assistant Director for the
Directorate of Engineering (ENG) in March 2013. In this position,
Khargonekar leads the ENG Directorate with an annual budget
of more than 890$ million. The ENG Directorate invests in
frontier engineering research and education, cultivates an
innovation ecosystem, and develops the next-generation of
engineers. He is a member of the senior leadership team at
NSF and thereby involved in setting priorities and policies at
NSF.
Khargonekar received B. Tech. Degree in electrical engineering
from the Indian Institute of Technology, Bombay, India, in
1977, and M.S. degree in mathematics and Ph.D. degree in
electrical engineering from the University of Florida in 1980
and 1981, respectively. He has held faculty positions at
the University of Florida, University of Minnesota, and The
University of Michigan. He was Chairman of the Department
of Electrical Engineering and Computer Science from 1997 to
2001 and also held the position of Claude E. Shannon Professor
of Engineering Science at The University of Michigan. From
2001 to 2009, he was Dean of the College of Engineering at
the University of Florida and is currently Eckis Professor of
Electrical and Computer Engineering there. He also served
briefly as Deputy Director of Technology at ARPA-E, U. S.
Department of Energy in 13-2012.
Human-Machine Networks and Intelligent Infrastructure
Khargonekar’s research and teaching interests are centered on
theory and applications of systems and control. His early work
was on mathematical control theory, specifically focusing on
robust control analysis and design. During the 1990’s, he was
involved in a major multidisciplinary project on applications
of control and estimation techniques to semiconductor
manufacturing. His current research and teaching interests
include systems and control theory, machine learning, and
applications to smart electric grid and neural engineering. He
has authored more than 130 refereed journal publications and
150 conference publications. He has supervised 32 doctoral
students. He has been recognized as a Web of Science Highly
Cited Researcher. He is a recipient of the NSF Presidential Young
Investigator Award, the American Automatic Control Council’s
Donald Eckman Award, the Japan Society for Promotion of
Science fellowships, the IEEE W. R. G. Baker Prize Award, the
IEEE CSS George Axelby Best Paper Award, the Hugo Schuck
ACC Best Paper Award, and the Distinguished Alumnus and
Distinguished Service Awards from the Indian Institute of
Technology, Bombay. He is a Fellow of IEEE and IFAC. At the
University of Michigan, he received the Arthur F. Thurnau
Professorship. In the past, he has served as Associate Editor
for IEEE Transactions on Automatic Control, SIAM Journal of
Control, Systems and Control Letters, and International J. of
Robust and Nonlinear Control.
7
DYNAMIC CO-EVOLVING
SOCIOTECHNICAL NETWORKS
PROF. JOHN
BARAS
Lockheed Martin Chair
in Systems Engineering,
University of Maryland
Abstract: We are in the age of networks and networked systems and infrastructures:
energy, communication, transportation, economic, biological, healthcare,
educational, human, social, web-based. This evolution and reality have created
unprecedented advances and are impacting every aspect of life and work. Several
of the emerging sociotechnical networks are interacting dynamically. Several key
and fundamental challenges have emerged including distributed autonomy,
human collaborative cognition and decision making, modeling and architectures.
We address these challenges as problems in cooperative multi-agent systems. We
describe a general model that involves several interacting dynamic multigraphs
and identify three fundamental research challenges underlying these systems.
We show that the framework of constrained coalitional network games captures
in a fundamental way the basic tradeoff of benefits vs. cost of collaboration, in
multi-agent systems, and demonstrate that it can explain network formation and
the emergence or not of collaboration. We analyze multi-metric problems in such
networks via a novel multiple partially ordered semirings approach. We apply these
methods to investigate the complex and polymorphic subject of trust in these
distributed systems including reputation and trust establishment, trust and mistrust
dynamics as well as “composite trust”. We demonstrate the fundamental role of trust
in collaboration and social networks. We close by describing novel and non-classical
stochastic network models that address successfully several known paradoxes in
understanding human collaborative cognition and decision making, within such coevolving sociotechnical networks with partial and constrained information.
BIOGRAPHY
Diploma in Electrical and Mechanical Engineering from
the National Technical University of Athens, Greece, 1970;
M.S., Ph.D. in Applied Mathematics from Harvard University
1973 ,1971. Since 1973, faculty member in the Electrical
and Computer Engineering Department, and in the Applied
Mathematics, Statistics and Scientific Computation Program,
at the University of Maryland College Park. Since 2000, faculty
member in the Fischell Department of Bioengineering.
Since 2014, faculty member in the Mechanical Engineering
Department. Founding Director of the Institute for Systems
Research (ISR), 1985 to 1991.
the Royal Swedish Academy of Engineering Sciences (IVA).
Received the 1980 George Axelby Prize from the IEEE Control
Systems Society, the 2006 Leonard Abraham Prize from the
IEEE Communications Society, the 2014 Tage Erlander Guest
Professorship from the Swedish Research Council, and a three
year (2017-2014) Senior Hans Fischer Fellowship from the
Institute for Advanced Study of the Technical University of
Munich, Germany. Professor Baras› research interests include
systems and control, optimization, communication networks,
signal processing and understanding, robotics, computing
systems and networks, network security and trust and modelbased systems engineering.
Since 1991, Founding Director of the Maryland Center for
Hybrid Networks (HYNET). Since 2013, Guest Professor at
the Royal Institute of Technology (KTH), Sweden. IEEE Life
Fellow, SIAM Fellow, AAAS Fellow, and a Foreign Member of
8
KAUST Conference, October 2015
HUMAN-MACHINE INTERACTION:
RATIONAL AND PREDICTABLE
FROM WHAT PERSPECTIVE?
PROF. AMY
PRITCHETT
David S. Lewis Associate
Professor of Cognitive
Engineering, Georgia Institute
of Technology
Abstract: This presentation will review common problems associated with humanmachine interaction in complex work domains, such as the aircraft flight deck. I will
then argue that human-interactive-systems should not assume a particular form of
interaction referenced to the automation’s functioning. Instead, the presentation
will discuss how the human expert is predictable from the perspective of the work
processes that s/he is attempting to perform. From this perspective, new methods
and computational models for designing effective human-machine function
allocation and interaction can be designed that addresses the humans’ needs
from the technology, and that supports the combined human-machine system
performance.
BIOGRAPHY
Dr. Pritchett’s research examines safety in dynamic multiagent environments. Her current studies particularly focus
on effective distribution of authority between agents in
general and air-ground and human-automation interaction
in particular, particularly in civil aviation. She also specializes
in translating these methods into safety analyses and in
identifying implications for air traffic operations, training, and
regulatory standards.
Dr. Pritchett has led numerous research projects sponsored
by industry, NASA and the FAA. She has also served via IPA
as Director of NASA’s Aviation Safety Program, responsible
for planning and execution of the program and serving on
several executive committees, including the OSTP Aeronautic
Science and Technology Sub-committee, and the executive
committees of CAST and ASIAS. She has published over
170 scholarly publications in conference proceedings and
Human-Machine Networks and Intelligent Infrastructure
scholarly journals. She has also won the RTCA William H.
Jackson Award and, as part of CAST, the Collier Trophy, and the
AIAA has named a scholarship for her and selected her for the
Lawerence Sperry Award. She served until 2014 as member
of the FAA REDAC and chaired the Human Factors REDAC
sub-committee. She has also served on numerous National
Research Council (NRC) Aeronautics and Space Engineering
Board and numerous NRC committees, including chairing
the recent committee examining FAA Air Traffic Controller
Staffing. She is editor-in-chief of the Human Factors and
Ergonomics Society’s Journal of Cognitive Engineering and
Decision Making. She is also a licensed pilot in airplanes and
sailplanes.
9
INVITED TALKS
WELCOME AND OPENING REMARKS
BIOGRAPHY
DR. TEOFILO
(JUN) ABRAJANO Dr. Teofilo (Jun) Abrajano is the director of the Office of Competitive Research Funds
Director, Office of Sponsored
Research, KAUST
(OCRF) at KAUST. He came to KAUST from the U.S. National Science Foundation,
where he held various positions including program director of Geobiology and
Low Temperature Geochemistry, head of Surface Earth Processes Section, division
director of Earth Sciences, and deputy assistant director of the GEO Directorate.
Prior to joining NSF, he was a professor of Earth and Environmental Sciences and the
director of the Environmental Sciences Program at Rensselaer Polytechnic Institute
in New York. His research focused on analytical developments in continuous
flow-stable isotope mass spectrometry and elucidating the isotope systematics,
biogeochemistry and geomicrobiology of modern and ancient aquatic systems.
Dr. Abrajano received his Ph.D. in Earth and Planetary Sciences from Washington
University St. Louis, Missouri in 1984. He is a Fulbright Fellow and a Fellow of the
Geological Society of America.
Human-Machine Networks and Intelligent Infrastructure
11
INTELLIGENT OPERATION OF LOW
EMISSIONS POWER GENERATION
PROF. ALI ABBAS
Laboratory for Multiscale
Systems, School of
Chemical and Biomolecular
Engineering
The University of Sydney,
Sydney NSW 2006, Australia
Abstract: Carbon-based fossil fuel resources comprise today more than %80 of global
primary energy consumption and they are a major contributor to carbon emissions
and arguably to climate change. Electricity generation from fossil power plants
(coal, oil, gas) is a major source of CO2 emissions and this makes them of significant
immediate interest. Integration of renewable solar thermal heat and/or power
with carbon capture-ready power plants offers a promising capability to bridge to
a sustainable future energy supply. However, the integrated operation is complex
and not immediately obvious to the human operator. Asc./Prof. Abbas shares in this
presentation his recent research and development on intelligent algorithms for the
operation of low-carbon power generation plants which extract maximum financial
benefits hidden in the dynamics of electricity load, electricity price, carbon price
and weather trends. The focus will be on a real-time control-optimization framework
developed across multiple time scales and which features high sample frequency
commensurate with electricity dispatch and control instrumentation levels but
which also serves longer term planning and scheduling. This framework is proposed
as a decision support HMI for flexible operation of carbon capture plants and for
implementation under contemporary emissions trading schemes. Such intelligent
algorithmic operation translates in enhanced techno-economics of electricity
supply while ensuring carbon emissions are reduced.
BIOGRAPHY
Dr Ali Abbas is Associate Professor at the School of Chemical
and Biomolecular Engineering at The University of Sydney
and is Founder and Director of the Laboratory for Multiscale
Systems. He conducts research in the area of Process Systems
Engineering with a primary focus on energy systems and
bio-systems. Dr Abbas’s ongoing model-based research in
carbon capture is helping power generators assess solutions
for integration of carbon capture and renewable energy.
He was awarded the Australia-Harvard Fellowship in 2011
and was recently nominated by the Australian Academy of
Technological Sciences and Engineering (ATSE) as a Future
Leader in low emissions clean coal technologies in 2012.
12
KAUST Conference, October 2015
APPLIED ROBOTICS IN THE ENERGY
SECTOR
MR. FADEL
ABDELLATIF
Saudi Aramco
Abstract: The field of applied robotic systems has grown tremendously in the
last decade. Robots are gradually gaining popularity and replacing humans in
performing tasks that are repetitive, hazardous or laborious. Advancements in
rapid prototyping technologies and 3D printing, along with easier accessibility to
embedded systems, have made robot development more approachable than ever.
Saudi Aramco’s Intelligent Systems Lab is pushing the envelope by developing
robotic solutions to challenges in the energy sector. Activities such as inspection and
maintenance of operation assets in various onshore and offshore environments are
among the tasks which can be effectively automated or assisted by robotic systems.
BIOGRAPHY
Fadl completed his Bachelor’s degree in Mechatronics
Engineering with distinction from University of Jordan in 2009.
He completed his Master’s degree in Mechanical Engineering
with distinction from KAUST in 2011. His research focus areas
include robotics, automation, systems modeling and control.
Fadl joined Saudi Aramco in 2012 and works in the Intelligent
Systems Lab in Saudi Aramco›s R&D Center at KAUST, where
he works on the development of robotic systems for the oil
and gas industry. He is a co-inventor in a number of patents in
applied robotics and he has received many awards including
the Saudi Aramco CEO Recognition for Excellence and the IFIA
Industrial Glory Medal.
Human-Machine Networks and Intelligent Infrastructure
13
DEMAND RESPONSE: A VALUABLE
ASSET FOR A SMARTER ELECTRIC
GRID
PROF. ALI
ALAWAMI
King Fahd University of
Petroleum and Minerals
(KFUPM)
Abstract: To maintain stable and reliable operation of the grid, electric power
generation and demand must match almost instantly. Traditionally, balanced
operation used to be performed by centralized control of the generation side so as
to «follow» the load because generators are readily accessible by the grid operator.
The smart grid can enable the grid operator to have access to the distributed electric
demand, either directly or through a third party. Hence, the demand may «respond»
to grid conditions, making generation-load balance, among other services, more
efficient than ever. In this talk, demand response (DR) as an asset for the grid
operator will be discussed. The talk will briefly cover the different services that
DR can provide to the grid, whether these services can be valued monetarily, and
whether they can be traded as commodities in an open-access electricity market.
New research directions in DR utilization will be presented. These include centralized
and distributed control, stochastic scheduling, and even some long-term planning
aspects of DR.
BIOGRAPHY
Ali Al-Awami obtained his Ph.D. in December 2010 from the
University of Washington, USA. Since then, he has been an
Assistant Professor at the Department of Electrical Engineering,
King Fahd University of Petroleum & Minerals, Dhahran, Saudi
Arabia. Ali›s research interests include the integration of
renewable energy sources, controllable demand, and electric
vehicles, into grid operations, deregulated electricity markets,
and smart grid technologies. He has a number of publications
in top journals and conferences in these fields. Dr. Al-Awami
won several awards from IEEE, University of Washington, and
KFUPM honoring the quality of his research. Dr. Al-Awami’s
industrial experience includes working for Saudi Electricity Co.
and Bonneville Power Administration.
14
KAUST Conference, October 2015
CITY DYNAMICS: A BIG DATA
ANALYTICS PLATFORM FOR CITIES
DR. ANAS
ALFARIS
King Abdelaziz City for
Science and Technology
(KACST) & Massachusetts
Institute of Technology
Abstract: Many cities in the world are experiencing rapid growth due to various
factors such as natural population growth, urbanization, and immigration. These
driving forces impact cities in myriad ways. By combining high resolution spatiotemporal data from multiple sources with the topological characteristics of the
city, as well as employing a wide range of methods for modeling and data analysis,
observed human behavior such as movement patterns and social interactions
can be evaluated. Furthermore, their impact on the city in a variety of domains
such as transportation, land use, energy consumption and ICT infrastructure can
be assessed. The City Dynamics platform offers the ability to simulate, query, and
aggregate models, while a new generation of web technologies provides powerful
visualization of geospatial and temporal big data. The City of Riyadh is used as a
case study for the City Dynamics Platform. This has helped shed light on the causes
of congestion affecting Riyadh’s road network, the mobility habits of the city’s
residents, the influence of commercial points of interest on urban movement
patterns. The results of the research can help policymakers address the challenges
of serving the current population and planning for the future by equipping them
with a level of insight that was not attainable before the era of big data.
BIOGRAPHY
Anas Alfaris is currently the Director of the Joint Centers of
Excellence Program (JCEP) at KACST which oversees several
joint centers with academia such as MIT, Stanford, Cambridge
and Oxford, as well as joint centers with industry such as IBM
and Boeing among others. He is also the co-director of the
Center for Complex Engineering Systems (CCES) at KACST and
MIT as well as the Co-director of the Decision Support Center
(DSC) of KACST and Boeing.
In addition, Alfaris is an Assistant Professor at the King Abdulaziz
City for Science and Technology (KACST) and visiting Assistant
Professor at the Massachusetts Institute of Technology (MIT).
Furthermore, Alfaris is the head of the consultative team for
the second National Science, Technology and Innovation Plan
for the Kingdom of Saudi Arabia. He also was the Chair of the
Task Force handling the National Strategy for Transforming
Saudi Arabia into a Knowledge Based Society. He served as
advisor to the Saudi Minister of Economy and Planning and is
also a member of the Steering Committee of the Saudi Arabia
Advanced Research Alliance (SAARA).
Human-Machine Networks and Intelligent Infrastructure
Alfaris’ research experience spans several fields including
Strategic Engineering, Network Analysis, Integrated Modeling
and Simulation, Multi-Disciplinary Analysis and Optimization
as well as Decision Support Systems. His current work focuses
on the development of Integrated Simulations for the planning
of large scale complex urban and Infrastructure systems.
Alfaris received his training in several disciplines including
civil architecture, engineering and computer science. He
received from MIT a Master of Science in Computation for
Design and Optimization from the Center of Computational
Engineering at the School of Engineering as well as a Ph.D. in
Design Computation.
15
RESILIENT MONITORING AND
CONTROL OF DISTRIBUTION
NETWORKS
PROF. SAURABH
AMIN
Massachusetts Institute of
Technology
Abstract: This talk will focus on the analysis of attacker-defender interactions on
distribution networks (DNs) using game-theoretic tools. Three attack models will be
considered: (i) strategic disruption of network links; (ii) manipulation of distributed
generation (DG) nodes; and (iii) stealing by fraudulent customers. In the first model,
the defender chooses flow routing strategies to maximize the expected effective
flow between source-destination pairs under strategic link disruptions. In the
second model, the defender responds to the adversary’s action by imposing DG
control and partial load shedding. In the third model, the DN operator (defender)
chooses the level of investment in technology to improve the detection and
identification of fraudulent customers. Full characterization of equilibrium strategies
will be presented for each model of attacker-defender interaction. The equilibrium
strategies provide practical recommendations for improving the resilience of
network monitoring and control tools in the face of strategic attacks.
BIOGRAPHY
Saurabh Amin is Robert N. Noyce Career Development Assistant
Professor in the Department of Civil and Environmental
Engineering, Massachusetts Institute of Technology (MIT).
His research focuses on the design and implementation of
high confidence network control algorithms for infrastructure
systems. He works on robust diagnostics and control problems
that involve using networked systems to facilitate the
monitoring and control of large-scale critical infrastructures,
including transportation, water, and energy distribution
systems. He also studies the effect of security attacks and
random faults on the survivability of networked systems,
and designs incentive-compatible control mechanisms to
reduce network risks. Dr. Amin received his Ph.D. in Systems
Engineering from the University of California, Berkeley in 2011.
He is a recipient of NSF CAREER award, and Google Faculty
Research award.
16
KAUST Conference, October 2015
DECENTRALIZED Q-LEARNING FOR
STOCHASTIC DYNAMIC GAMES
PROF. GURDAL
ARSLAN
University of Hawaii
Abstract: There are only a few learning algorithms applicable to stochastic dynamic
games. Learning in games is generally difficult because of the non-stationary
environment in which each decision maker aims to learn its optimal decisions with
minimal information in the presence of the other decision makers who are also
learning. In the case of dynamic games, learning is more challenging because, while
learning, the decision makers alter the state of the system and hence the future
cost. In this paper, we present decentralized Q-learning algorithms for stochastic
dynamic games, and study their convergence for the weakly acyclic case. We show
that the decision makers employing these algorithms would eventually be using
equilibrium policies almost surely in large classes of stochastic dynamic games.
BIOGRAPHY
Gürdal Arslan received Ph.D. degree in electrical engineering
from the University of Illinois at Urbana-Champaign, in 2001.
From 2001 to 2004, he was an Assistant Researcher in the
Department of Mechanical and Aerospace Engineering,
University of California, Los Angeles. In August 2004, he joined
the Electrical Engineering Department at the University of
Hawaii, Manoa.
His current research interests lie in the design of cooperative
(multi-agent) systems using game theoretic methods. Recent
applications of his research include autonomous resource
allocation for mission planning, multi-sensor deployment,
traffic management, and cooperative multi-user MIMO
signaling in wireless communication systems.
He is a member of the IEEE Control Systems Society and he
received the National Science Foundation CAREER Award on
«Cooperative Systems Design - Stochastic Games Approach»
in May 2006.
Human-Machine Networks and Intelligent Infrastructure
17
ON AUTONOMIC COMPUTING
PROF. MARIETTE
AWAD
American University of Beirut
Abstract: Inspired by biology, autonomic systems (AS) are the building blocks of
pervasive computing and the focus of many industry leaders including IBM and
Microsoft who are researching different components of AS. Similarly to the nervous
system, AS are expected to be self-healing, self-configured, self-protected and selfmanaged in their environment, all of which represent a major challenging problem
for research to yet solve in what resemble a science fiction. In an attempt to shed
light to this evolving field, this talk starts first with an overview of the definitions
and key elements of AS before providing the audience with latest achievement in
this field that looks like science fiction. Finally, AS related research happening at the
American University of Beirut will conclude the talk.
BIOGRAPHY
Prof. Mariette Awad is an associate professor in the Electrical and
Computer Engineering Department of the American University of
Beirut. She received her PhD in electrical engineering in 2007 and she
has been a visiting professor at Virginia Commonwealth University,
Intel Mobile Group, and MIT. She was invited by CVC labs, Google
and Qualcomm to present her work on machine learning and image
processing. She has published in numerous conferences and journals
and is managing few grants. She is an active member in IEEE and
reviewer for IEEE transactions. Prior to her academic position, she
was with IBM System and Technology group in Vermont as a wireless
product engineer. Over the years, her technical leadership and
innovative spirit has earned her management recognition, several
business awards, and multiple patents at IBM. Her current research
interests include machine learning, game theory, energy aware
computing and cybernetics.
18
KAUST Conference, October 2015
ALGORITHMS FOR AN INTELLIGENT
AIR TRANSPORTATION
INFRASTRUCTURE
PROF. HAMSA
BALAKRISHNAN
Massachusetts Institute of
Technology
Abstract: The air transportation system involves significant interactions between
automation and human decision-makers. A key challenge in designing and operating
an intelligent air transportation infrastructure lies in reliably predicting the decisions
made by the human operators, and responding appropriately. Modeling these
systems and providing decision support to operators needs an understanding of
the objective functions in the decision processes, and the development of efficient
algorithms that can optimize them. In this talk, I will discuss these problems in
the context of data-driven modeling of air traffic controller utility functions, and
resource allocation in multi-stakeholder environments.
BIOGRAPHY
Hamsa Balakrishnan is an Associate Professor of Aeronautics
and Astronautics at the Massachusetts Institute of Technology
(MIT). Before joining MIT, she was at the NASA Ames Research
Center, after receiving her PhD from Stanford University and
a B.Tech. from the Indian Institute of Technology Madras.
Her research is in the design, analysis, and implementation
of control and optimization algorithms for large-scale cyberphysical infrastructures, with an emphasis on air transportation
systems. Her contributions include airport congestion control
algorithms, air traffic routing and airspace resource allocation
methods, machine learning for weather forecasts and flight
delay prediction, and methods to mitigate environmental
impacts.
She was a recipient of the NSF CAREER Award in 2008, the Kevin
Corker Award for Best Paper of ATM2011-, the inaugural CNA
Award for Operational Analysis in 2012, the AIAA Lawrence
Sperry Award in 2012, and the American Automatic Control
Council›s Donald P. Eckman Award in 2014.
Human-Machine Networks and Intelligent Infrastructure
19
DECENTRALIZED AND OPTIMAL
CONTROL OF INTER-AREA
OSCILLATIONS IN POWER
NETWORKS
PROF. FLORIAN
DORFLER
Swiss Federal Institute of
Technology (ETH)
Abstract: Local and inter-area oscillations in bulk power systems are typically
identified using spatial profiles of poorly damped modes, and they are mitigated
via carefully tuned decentralized controllers. Recent research efforts have been
aimed at developing wide-area control strategies that involve communication of
remote signals. In conventional wide-area control, the control structure is fixed
a priori typically based on modal criteria, and structured controllers are tuned
accordingly. In this seminar, we employ non-modal tools to analyze and control
inter-area oscillations. Our input-output analysis examines power spectral density
and variance amplification of stochastically forced systems and offers new insights
relative to modal approaches. To improve upon the limitations of conventional widearea control strategies, we also study the problem of signal selection and optimal
design of sparse and block-sparse wide-area controllers using the the recentlyintroduced paradigm of sparsity- promoting optimal control. In our design, we
preserve rotational symmetry of the power system by allowing only relative angle
measurements in the distributed controllers. We use the New England and New York
power grid models to examine performance tradeoffs and robustness of different
control architectures. We demonstrate that optimal and fully-decentralized control
strategies can effectively guard against local and inter-area oscillations.
BIOGRAPHY
Florian Dörfler is an Assistant Professor at the Automatic
Control Laboratory at ETH Zürich. He received his Ph.D. degree
in Mechanical Engineering from the University of California at
Santa Barbara in 2013, and a Diplom degree in Engineering
Cybernetics from the University of Stuttgart in 2008. From
2013 to 2014 he was an Assistant Professor at the University
of California Los Angeles. His primary research interests are
centered around distributed control, complex networks, and
cyber–physical systems currently with applications in energy
systems and smart grids. He is a recipient of the 2009 Regents
Special International Fellowship, the 2011 Peter J. Frenkel
Foundation Fellowship, the 2010 ACC Student Best Paper
Award, the 2011 O. Hugo Schuck Best Paper Award, and the
2014-2012 Automatica Best Paper Award. As a co-advisor
and a co-author, he has been a finalist for the ECC 2013 Best
Student Paper Award.
20
KAUST Conference, October 2015
THE INTERNET IS NOT AN
INTELLIGENT INFRASTRUCTURE:
CAUSES AND POTENTIAL
SOLUTIONS
PROF.
CONSTANTINE
DOVROLIS
Abstract: The Internet is one of the most complex artifacts humans have ever created. Its
Georgia Institute of
Technology
designers. These mechanisms have reached such a high level of autonomy and complexity
success is largely due to sophisticated mechanisms that automatically perform functions
such as routing, congestion control, extremely fast packet forwarding, traffic engineering,
etc, with minimal human intervention. As a result, the Internet can automatically react to
failures and traffic perturbations, it can grow in an organic and decentralized manner, and
it can support new applications and services that were not even dreamed of by its original
that one can arguably think of the Internet as a clear instance of «intelligent infrastructure.»
What is not widely known, however, is that there is a critical component of the Internet
that is still %100 «manual», meaning that it is driven exclusively by human decisions
and actions. This component is the peering agreements and interconnections between
the 50,000 or so Autonomous Systems (ASes) that form the Internet. The Internet is an
ecosystem of about 50,000 Autonomous Systems (or ASes) that operate independently,
having different objectives and operational constraints. What glues the Internet together is
the bilateral techno-economic agreements that form the interconnections between these
ASes. The decision to interconnect two ASes is a slow and often difficult process because
the representatives of the two organizations have to agree on both technical and economic
aspects of the interconnection. These interconnections (transit, peering, paid-peering, etc)
have evolved over the last 20 years or so, since the commercialization of the Internet in the
mid-nineties, in a rather ad-hoc manner, often resulting in bilateral or multilateral disputes
about who should peer with whom, whether one of the two parties should pay the other,
and about the conditions that an interconnection should satisfy (e.g., balanced transit ratios).
These problems result in congested interconnections and, in some cases, unreachability
problems that can affect millions of Internet users. In this talk, I will make the case that the
Internet cannot serve as a truly «intelligent infrastructure» given the slow and ad-hoc manner
in which peering interconnections are performed. Instead, I will describe a novel peering
framework (work-in-progress) in which these peering interconnections are performed
based on sophisticated algorithms that set the price for individual routes and network paths
dynamically and autonomously.
BIOGRAPHY
Dr. Constantine Dovrolis is a Professor at the College of
Computing of the Georgia Institute of Technology. He
received the Computer Engineering degree from the Technical
University of Crete in 1995, the M.S. degree from the University
of Rochester in 1996, and the Ph.D. degree from the University
of Wisconsin-Madison in 2001. He has held visiting positions at
Thomson Research in Paris, Simula Research in Oslo, FORTH in
Crete, and University of Thessaly in Volos. His current research
focuses on cross-disciplinary applications of network science
Human-Machine Networks and Intelligent Infrastructure
in biology, climate science and neuroscience. He has also
worked on the evolution of the Internet, Internet economics,
and on applications of network measurement. He received the
National Science Foundation CAREER Award in 2003.
21
UTILITIES OF THE FUTURE:
NEW BUSINESS MODELS AND
REGULATORY FRAMEWORKS
DR. ROLANDO
FUENTES
King Abdullah Petroleum
Studies and Research Center
(KAPSARC)
Abstract: KAPSARC is engaged in exploratory modeling and analysis of electric
power distribution systems. Existing power system arrangements need to be
revisited in light of advances in distributed energy resources (DER) and smart grid
technologies. Understanding the institutional arrangements, as well as behavioral
responses and operational profiles of the technologies is critical for integrating DER
into the industry’s existing framework. New institutional designs, whether they are
business or regulatory, may be needed to ensure the provision of electric power
remains clean, affordable and reliable.
BIOGRAPHY
Dr. Rolando Fuentes is a Research Fellow at the King Abdullah
Petroleum Studies and Research Center (KAPSARC). He holds an
MSc in Environmental and Resource Economics from University
College London and a PhD from the London School of Economics
(LSE) for research into the power sector reform in Mexico. Dr. Fuentes
was Fellow of the LSE, where he lectured and taught courses in
Environmental Impact Assessment and Environmental Policy. He
has also been an associate of the Oxford Institute of Energy Studies
and IHS-CERA. He was Director of Project Programming in the
Mexican Ministry of Energy after holding the position of Director of
International Negotiations at the same Ministry.
22
KAUST Conference, October 2015
LAYERED CONTROL OF ROAD
FREIGHT TRANSPORT
PROF. KARL
JOHANSSON
Royal Institute of Technology
(KTH)
Abstract: Freight transportation is of outmost importance for our society and is
continuously increasing, particularly in the emerging economies. At the same time,
transporting goods on roads accounts for about %26 of all energy consumption
and %18 of greenhouse gas emissions in the European Union, with similar figures
elsewhere. Despite the influence the transportation system has on our energy
consumption and the environment, road transportation is mainly done by individual
long-haulage trucks with no real-time coordination or global optimization. In this
talk, we discuss how modern information and communication technology supports
a cyber-physical transportation system architecture with an integrated logistic
system coordinating fleets of trucks traveling together in vehicle platoons. From
the reduced air drag, platooning trucks traveling close together can save more than
%10 of their fuel consumption. Control and estimation challenges and solutions
on various level of this transportation system will be presented. It will be argued
that a system architecture utilizing vehicle-to-vehicle and vehicle-to-infrastructure
communication enable receding horizon optimal control of individual trucks as well
as optimized platoons and fleets of platoons. Experiments done on test trucks Will
illustrate system performance and safety requirements. Some preliminary results
from a large-scale evaluation currently being performed on the highway road
network in Northern Europe will also be discussed. The presentation will be based
on joint work with collaborators at KTH and at the truck manufacturer Scania.
BIOGRAPHY
Karl H. Johansson is Director of the ACCESS Linnaeus Centre and
Journal of Control. He has been Guest Editor for a special issue of IEEE
Professor at the School of Electrical Engineering, KTH Royal Institute of
Transactions on Automatic Control on cyber-physical systems and
Technology, Sweden. He is a Wallenberg Scholar and has held a Senior
one of IEEE Control Systems Magazine on cyber-physical security. He
Researcher Position with the Swedish Research Council. He also heads
was the General Chair of the ACM/IEEE Cyber-Physical Systems Week
the Stockholm Strategic Research Area ICT The Next Generation. He
2010 in Stockholm and IPC Chair of many conferences. He has served
received MSc and PhD degrees in Electrical Engineering from Lund
on the Executive Committees of several European research projects
University. He has held visiting positions at UC Berkeley, California
in the area of networked embedded systems. He received the Best
Institute of Technology, Nanyang Technological University, and
Paper Award of the IEEE International Conference on Mobile Ad-hoc
Institute of Advanced Studies Hong Kong University of Science
and Sensor Systems in 2009 and the Best Theory Paper Award of the
and Technology. His research interests are in networked control
World Congress on Intelligent Control and Automation in 2014.
systems, cyber-physical systems, and applications in transportation,
In 2009 he was awarded Wallenberg Scholar, as one of the first
energy, and automation systems. He has been a member of the IEEE
ten scholars from all sciences, by the Knut and Alice Wallenberg
Control Systems Society Board of Governors and the Chair of the
Foundation. He was awarded Future Research Leader from the
IFAC Technical Committee on Networked Systems. He has been on
Swedish Foundation for Strategic Research in 2005. He received the
the Editorial Boards of several journals, including Automatica, IEEE
triennial Young Author Prize from IFAC in 1996 and the Peccei Award
Transactions on Automatic Control, and IET Control Theory and
from the International Institute of System Analysis, Austria, in 1993.
Applications. He is currently a Senior Editor of IEEE Transactions
He received Young Researcher Awards from Scania in 1996 and from
on Control of Network Systems and Associate Editor of European
Ericsson in 1998 and 1999. He is a Fellow of the IEEE.
Human-Machine Networks and Intelligent Infrastructure
23
THE SMART OCEAN:
TRANSFORMATIVE INTEGRATED
OCEAN OBSERVATION NETWORKS
PROF. BURT
JONES
Professor of Marine Science
KAUST
Abstract: Recent advances in robotic vehicle technology have transformed our
ability to observe the ocean at a range of scales and for a range of ocean-related
questions and problems, and to sustain these observations for extended time
periods. Modern ocean modeling has developed sophisticated capabilities for
assimilating data streams from these vehicles as well as more traditional fixed
platforms to provide increasingly accurate projections (nowcasts and forecasts) of
the ocean state for physical and biogeochemical processes within the ocean. Yet the
processes of observing and modeling the ocean continue to function as somewhat
independent entities, with the observations focused on providing observational
scientists with results that can be readily evaluated analytically and/or statistically,
while simultaneously providing input into real-time -4D models for nowcasts and
forecasts of marine physical dynamics. In the next generation of ocean observing
and modeling, the two systems come together as a single integrated observing and
forecasting tool that redefines how we observe and interpret the oceans and seas.
In this approach, observations in the fully coupled observation-modeling system
are tuned by the model through dynamic repositioning of robotic vehicles and
sensor control to improve the validity of forecasts through reduction of uncertainty,
increased resolution of specific phenomena or features, and to enhance our ability
to image and visualize the -4dimensional ocean.
BIOGRAPHY
Prof. Burton Jones is a professor in the Biological and
Environmental Sciences and Engineering Division and a
member of the Red Sea Research Center at KAUST. He joined
KAUST in March 2012.
He is a member of the American Geophysical Union, the
Association for the Sciences of Limnology and Oceanography,
and The Oceanography Society.
He received his bachelor of science degree in biological
engineering from Rose-Hulman Institute of Technology and his
doctorate in biological oceanography from Duke University.
After a postdoctoral fellowship at Bigelow Laboratory, he
joined the faculty at the University of Southern California.
Prof. Jones has been involved in implementing a regional
coastal observing system for Southern California in the United
States. While in Southern California, he was involved in the
development of collaborations between academic research,
regional public and private agencies, and he co-chaired the
Executive Steering Committee of Southern California Coastal
Ocean Observing System. He has participated in numerous
panels for developing policy for urban uses (including sewage,
runoff, and desalination brine discharges) of the coastal ocean.
24
KAUST Conference, October 2015
TOWARDS OIL AND GAS
INFRASTRUCTURE SUSTAINABILITY
ONSHORE AND OFFSHORE
PROF. MANSOUR
KARKOUB
Texas A&M University at
Qatar
Abstract: The petroleum industry is one of the largest in the world. It was estimated
in 2013 that the world uses more than ninety million barrels of oil and ten billion
cubic meters of natural gas per day. The production cost of oil is one of the reasons
for the high oil prices, which includes the cost of downtime due to failures in drilling,
transportation means, and maintenance. To increase the efficiency and production
of oil, researchers are constantly trying to develop ways to reduce non-productive
time (NPT), which is time spent fixing problems in (1) drilling such as stuck pipes,
reaming, fishing operations or unscheduled casing, and (2) pipelines such as leaking
pipelines, stuck pigs, corrosion, and displaced pipelines. For drilling, there are many
factors involved that reduce the efficiency of drilling. One of the largest limiting
factors in drilling performance is vibration. Controlling and suppressing the vibration
of the drillstring provides benefits to drilling performance such as better borehole
quality, reduction in NPT, greater tool reliability and overall lower drilling costs. In
order to make the necessary modifications to the BHA and drillstring, the drillstring
dynamics must first be understood and modeled. Here, we will talk about the design
of a smart drill rig to be used in identifying the influencing factors of drillstring failure.
Another costly, but necessary operation performed by the Oil and Gas industry is
pipeline inspection. The Oil and Gas industry operates and maintains thousands
of miles of pipelines consisting of steel and cast iron pipes to transport crude and
natural gas throughout the world. The steel Oil and Gas mains are prone to time
dependent defects such as corrosion, cracks, and dents which can reduce safety and
security of service and threaten the environment if failure occurs. Inspection plays a
key role in the integrity management of pipelines. Millions of miles of pipelines are
operated above the ground and on the seabed, which require period inspection to
avoid catastrophic failures. Here, we will talk about the use of smart autonomous
robotic systems in the inspection and defect detection of pipelines.
BIOGRAPHY
Mansour Karkoub is a professor of mechanical engineering
at Texas A&M University at Qatar. He received the B.S. degree
in Mechanical Engineering with Highest Distinction in 1984,
M.S.M.E. in 1990, and PhD in 1994 all from the University of
Minnesota, Minneapolis, Minnesota, USA. In 2003, Dr. Karkoub
received an HDR (Habilitation a Diriger des Recherches) from
the University of Versailles, France. He held faculty positions at
the College of Petroleum and Engineering, Kuwait University
and the Petroleum Institute in the UAE and visiting professor
position at the French Research Institute in Informatics and
Automation (INRIA) from 2003-2002. He published more than
150 peer-reviewed articles and one textbook on computer
Human-Machine Networks and Intelligent Infrastructure
aided design. He is a fellow of the Institute of Engineering
Technology, a fellow of the Institute of Mechanical Engineering,
and a senior member of IEEE. His research interests are robust
control, robotics, mechatronics, and vibration engineering.
25
DISTRIBUTED LEARNING
DYNAMICS: CONVERGENCE IN
ROUTING GAMES AND BEYOND
MR. WALID
KRICHENE
University of California
Berkeley
Abstract: Routing games offer a simple yet powerful model of congestion in traffic
networks, both in transportation and communication systems. The congestion in
such systems is affected by the combined decision of the agents (drivers or routers),
so modeling the decision process of the agents is important, not only to estimate
and predict the behavior of the system, but also to be able to control it (e.g. by
imposing tolls). This decision process is often called learning, as agents «learn» about
the system (such as the cost functions) or about the other agents (their decisions).
We propose and study different models of learning with the following requirement:
the joint learning dynamics should converge to the Nash equilibrium of the game.
In particular, we focus on a few essential properties: Is the model robust to stochastic
perturbations (such as measurement noise)? Does the model allow heterogeneous
learning (different agents may follow different learning strategies)? Can we provide
guarantees on the convergence rate? We study these questions using tools from
online learning theory and stochastic approximation theory. We also briefly discuss
extensions to other classes of games and related problems.
BIOGRAPHY
Walid Krichene received his Bachelor›s degree in 2008, and his
Master›s degree in Applied Mathematics in 2010, both from the Ecole
des Mines Paristech. He is currently pursuing his Ph.D. in Electrical
Engineering and Computer Sciences at the University of California at
Berkeley, advised by Professor Alexandre Bayen. His research focuses
on learning dynamics in distributed systems, both deterministic and
stochastic, with human and computer agents. His previous research
work includes collaboration with the CDC department at Caltech,
the CAOR group at the Ecole des Mines, and Criteo labs. Walid has
received multiple awards including the Valedictorian Baccalaureate
Prize from the Tunisian president, a Bronze medal at the Pan African
Mathematics Olympiads (2005), two Outstanding Graduate Student
Instructor awards from UC Berkeley (2015 ,2014), and the Leon Chua
Award for outstanding achievement in nonlinear science (2015).
26
KAUST Conference, October 2015
MODELING, OPTIMIZATION
AND CONTROL OF SUSTAINABLE
DESALINATION PLANTS
PROF. MERIEM
LALEG
KAUST
Abstract: Water desalination is the primary source of clean water in several
countries including Saudi Arabia. The process of desalination is usually energy
inefficient. Therefore it is highly desirable to incorporate renewable energy in this
process. In this talk, I will focus on modeling, optimization and control of distributed
heat transfer mechanisms illustrated by the performance control of an emerging
sustainable water desalination technique called membrane distillation. Membrane
distillation is a thermal separation technique driven by a vapor pressure difference
across a hydrophobic membrane. It offers several advantages over conventional
desalination methods. Among these advantages, the membrane distillation water
desalination system can be integrated with renewable and waste energy sources. As
it is well known, the objective is always to operate any process at optimal settings,
which reduces the operational costs and guaranties the performance and stability
of the system. This is true for the solar-powered membrane distillation water
desalination, where the objective is to maximize the water production and reduce
the energy consumption at the same time. However, very limited work has been
done in terms of process control and optimization of this system. In this talk, I will
present two approaches for modeling the membrane distillation process then I will
introduce the control problem with some proposed solutions.
BIOGRAPHY
Taous Meriem Laleg-Kirati is an assistant professor in the division of
Computer, Electrical and Mathematical Sciences and Engineering at
KAUST. She joined KAUST in December 2010.
From 2009 to 2010, she was working as a permanent research
scientist at the French Institute for research in Computer Sciences
and Control Systems (INRIA) in Bordeaux. She earned her Ph.D in
Applied Mathematics from INRIA Paris, in 2008. She holds a Master
in control systems and signal processing from University Paris 11 in
France.
Her research interests include, modeling, estimation, and control of
physical systems, inverse problems, signal and image analysis. She
considers applications in engineering and biomedical fields.
Human-Machine Networks and Intelligent Infrastructure
27
CONNECTING DISTRIBUTED
CONTROL AND DISTRIBUTED
OPTIMIZATION IN THE POWER
GRID AND GENERAL NETWORK
SYSTEMS
PROF. NA LI
Harvard University
Abstract: Conventional operation of the power grid is hierarchical and divided
into two vertical layers at different time scales. At the top decision-making layer,
abstract decision-makers such as generator firms, utilities, and independent system
operators adopt optimization approaches to determine nominal economicallyefficient operating points at a slow-time scale. At the bottom dynamical-control
layer, physical dynamical systems such as power plants are regulated to achieve realtime supply-demand balance while maintaining the system around the nominal
operating points. With the increasing penetration of distributed energy resources,
this hierarchical structure induces large economic inefficiency due to the fast
and large fluctuations of the renewable generations. In this talk, I will present our
work on distributed economically-efficient control, which improves the economic
efficiency of the fast-time scale dynamical control. In the proposed control, the
optimization computation is implicitly carried out by the physical dynamics of the
control layer; in other words, the dynamics and control actions of physical plants are
explicitly incorporated into the optimization algorithms. The resulting dynamicalcontrol will automatically track the system efficient points regardless of uncertain
system disturbances. I will also introduce a framework, reverse-forward engineering,
which guides us the design of distributed economically-efficient control. Lastly, I will
extend this framework to general network systems.
BIOGRAPHY
Na Li is an assistant professor in the School of Engineering and
Applied Sciences in Harvard University since 2014. She received
her PhD degree in Control and Dynamical systems from California
Institute of Technology in 2013 and was a postdoctoral associate
of the Laboratory for Information and Decision Systems at
Massachusetts Institute of Technology. Her research lies in the design,
analysis, optimization and control of distributed network systems,
with particular applications to power networks and systems biology/
physiology. She entered the Best Student Paper Award finalist in the
2011 IEEE Conference on Decision and Control.
28
KAUST Conference, October 2015
DESIGN AND DEPLOYMENT OF
CONTROLLERS IN PASSIVELY
OBSERVED NETWORKS
PROF.
DONATELLO
MATERASSI
University of Tennessee
Abstract: Many systems can only be passively observed. By passive observations, it is
meant that the manifest variables are not the system response to known inputs that
have been actively injected to probe or identify the network. Rather the available
measurements are being acquired while the system is currently operating and
forced by potentially unknown excitations. The capability of designing controllers
in this kind of scenario is of paramount importance for any unknown or partially
known system fulfilling critical or uninterruptible functions (i.e., a power grid, a
logistic system) or in situations where it is impractical or too expensive to inject
known probing signals into the system (i.e., a gene network, a financial network).
Under this paradigm, a fundamental goal is to shape the behavior of the system
while it is performing critical operations by synthetizing controllers that could be
readily deployed in-line. The talk proposes a methodological paradigm to design
controllers for networked systems when the interconnection structure of the system
is uncertain and only passive observations are available. In the proposed framework,
given a networked system, it is possible to design a control algorithm that is at
the same time local (it processes measurements from a limited number of nodes);
topologically robust (it tolerates uncertainties in the network topology); noninterfering (it makes use only of passive observations); and proficient (it guarantees
adequate levels of stability, performance, and resilience).
BIOGRAPHY
Donatello Materassi holds a Laurea in «Ingegneria Informatica»
and a «Dottorato di Ricerca» in Electrical Engineering/
Nonlinear Dynamics and Complex Systems from Universita›
degli Studi di Firenze, Italy. He has been a research associate
at University of Minnesota (Twin Cities) till 2011, is a postdoctoral researcher at Laboratory for Information and Decision
Systems (LIDS) at the Massachusetts Institute of Technology
and a lecturer at Harvard University till 2014. Currently he is
an assistant professor at University of Tennessee in Knoxville.
His main research interests are graphical models, stochastic
systems and cybernetics
Human-Machine Networks and Intelligent Infrastructure
29
THE DYNAMICS OF SOCIAL
INFLUENCE
DR. BARY
PRADELSKI
Swiss Federal Institute of
Technology (ETH)
Abstract:Individual behaviors such as smoking, fashion, and the adoption of new
products is influenced by taking account of others› actions in one›s decisions. We
study social influence in a heterogeneous population and analyze the long-run
behavior of the dynamics. We distinguish between cases in which social influence
arises from responding to the number of current adopters, and cases in which social
influence arises from responding to the cumulative usage. We identify the equilibria
of the dynamics and show which equilibrium is observed in the long-run. We fi?nd
that the models exhibit diff?erent behavior and hence this diff?erentiation is of
importance. We also provide an intuition for the diff?erent outcomes.
BIOGRAPHY
Bary is a game theorist who is interested in distributed learning,
bounded rationality, and stochastic processes. To date his
work considered the problems of convergence, convergence
rate and equilibrium selection for different cooperate and
non-cooperative games. Bary studied Mathematics at the
TU Munich, Ecole Polytechnique Paris, and Oxford University
where he just finished his DPhil. He now joins the Department
of Computational Social Sciences at ETH Zurich as a postdoctoral researcher.
30
KAUST Conference, October 2015
DRONE VISION: APPLYING
COMPUTER VISION ALGORITHMS
TO A SWARM OF NETWORKED
AERIAL VEHICLES FOR REMOTE
SENSING AND UNDERSTANDING
DR. NEIL SMITH
FalconViz
Abstract: The miniaturization of sensors and reduction in cost of aerial vehicles
has opened a new avenue for computer vision algorithms to be applied to a swarm
of highly mobile autonomous aerial vehicles. This provides new challenges and
opportunities to adapt and improve upon computer vision algorithms originally
designed for terrestrial applications. Work in this field shows great promise for
developing a swarm of UAVs that can sense and react to dynamic environments
providing a greater level of autonomy that currently cannot be achieved solely
by use of other sensing devices such as GPS, IMUs, or sonar. The applications are
broad from urban reconstruction, autonomous navigation, object avoidance and
real-time tracking. In this talk, we will present our current research at KAUST with
UAVs including recent developments in dense urban reconstruction and distributed
vision-based aerial tracking.
BIOGRAPHY
Dr. Smith’s research interests focus on the merging of
computational science with a variety of domain driven research
topics including disciplines such as archaeology, cultural
heritage and urban planning. Prior to his current position as
Research Scientist at the Visual Computing Center in KAUST,
he was a postdoctoral research fellow at the California Institute
of Telecommunications and Information Technology (Calit2),
University of California, San Diego (UCSD). He played a critical
role in developing novel software and hardware applications
for archaeology and cultural heritage through Calit2’s
Center of Interdisciplinary Science of Art, Architecture and
Archaeology (CISA3). Since 1999, Dr. Smith has been involved
in the integration of geographic information systems (GIS),
database management systems (DBMS) and remote sensing
to archaeology and cultural heritage. During this period
both ArchField (field recording technique integrating Total
Stations and SfM) and ArtifactVis2 (a 3D stereoscopic scientific
Human-Machine Networks and Intelligent Infrastructure
visualization tool for archaeologists) were developed. Current
research has focused on novel computer vision solutions
that incorporate geographic information systems, scientific
visualization, informatics, and computer vision. In particular,
research at VCC has led to the development of aerial based
scanning and tracking solutions using unmanned multi-rotor
copters mounted with imaging sensors. Dr. Smith is also cofounder and CEO of FalconViz a KAUST spin-off company
focused on UAV based surveying and mapping for the
construction sector.
31
SECURITY BADGES
Required at many access points
GYM
Cardio/fitness/pool at Harbor Sports Club
(from 6 a.m. to 10 p.m.)
About a five-minute walk from the KAUST Inn
COFFEE
Coffee shops available at major locations and buildings:
Building 16, University Library, Discovery Square
(House of Donuts located at Building 2)
ATM
Available at SAMBA bank, Campus Diner and Tamimi Supermarket
32
KAUST Conference, October 2015
Human-Machine Networks and Intelligent Infrastructure
33
HUMAN-MACHINE NETWORKS AND
INTELLIGENT INFRASTRUCTURE
KAUST, 2015