We live in an interconnected society where human-to-human interactions
have been augmented by technical systems: social media has not only
grown the scale of our networks from tens to potentially millions of
nodes, it also mediates our interactions and merges them with AI agents
including recommender systems and bots. At the same time, we become
increasingly aware of large-scale sociological effects like
polarization, echo chambers, homophily and crowd effects. Understanding
the mechanisms by which such macroscopic behaviors emerge from
individual-level interactions, where opinions and beliefs propagate
dynamically over sociotechnical networks is the fundamental research
challenge in social network theory.  In this talk, I will give a broad
overview of our ELLIIT project group’s ongoing research in this area,
and explain how fairly traditional systems and control theory can be
used to model and understand emerging behaviors in social networks. I
will also highlight recent work that can model how individual bias
combined with network effects may lead to an evolving distrust in
centrist media among certain groups.

Emma Tegling is a Senior Lecturer with the Department of Automatic
Control at Lund University, Sweden. She received her Ph.D. degree in
Electrical Engineering from KTH Royal Institute of Technology,
Stockholm, Sweden in 2019, and her M.Sc. and B.Sc. degrees, both in
Engineering Physics, from the same institute in 2013 and 2011,
respectively. During 2019–2020 she was a Postdoctoral Research Fellow
with the Institute of Data, Systems, and Society (IDSS) at the
Massachusetts Institute of Technology (MIT), Cambridge, USA.  She has
also spent time as a visiting researcher at Caltech, the Johns Hopkins
University, and UC Santa Barbara. Emma’s research interests are within
analysis and control of large-scale networked systems. She has a
particular interest for control challenges in distributed electric power
networks and social networks.

Techniques from computer graphics are used in a wide variety of
applications ranging from photo-realistic rendering of products and
digital design to simulation of image and sensor data for machine
learning and computer vision applications including self-driving
vehicles and robotics. The simulation process entails building models
describing the light transport in a scene, scattering at surfaces and
the sensors themselves. This presentation will give an overview of the
work carried out in the Computer Graphics and Image Processing group at
the department for Science and Engineering at Linköping University, with
a focus towards how we improve accuracy and realism in graphics-based
simulation and how we reduce the computational costs and improve
rendering speed.

Professor Jonas Unger is leading the research efforts within Computer
Graphics and Image Processing group in the division for Media and
Information Technology at the department of Science and Technology. The
vision of the group is to research and develop new theory and technology
for computational imaging by fusing computer graphics,vision and sensors
with human perception and machine learning to capture, digitize and
reason about aspects of the world that have not been possible before.
With a strong foundation in the theoretically oriented research, the
group is active within a number of industrial and academic
collaborations directed towards development of state-of-the-art
applications ranging from 3D-reconstruction of scenes, photorealistic
image synthesis and digitization of optical material properties to
computer vision for heart surgery, AI driven diagnostics tools for
medical applications, perceptual display algorithms, and software for
autonomous systems such as self-driving cars and robot navigation.

After a short review of the concept of Information Freshness (a.k.a. Age of Information) where its role as a metric and a tool is developed, it will be shown how it leads naturally to other information properties that can be called Semantic. The term “Semantic” needs clarification and represents a paradigm shift in communications from the traditional one that has been based on Shannon’s Theory. Semantic properties are meaningful and useful in a variety of applications. They also provide a potential link between Information Theory and Signal Processing. Finally, the concept of freshness is linked to the current focus on data driven methodologies that emerge from the transition to Artificial Intelligence and Machine Learning.

Anthony Ephremides (etony@umd.edu) received his B.S. degree from the National Technical University of Athens (1967), and M.S. (1969) and Ph.D. (1971) degrees from Princeton University, all in Electrical Engineering. He has been at the University of Maryland since 1971, and held a joint appointment as Professor in the Electrical Engineering Department and a founding member of the Institute of Systems Research (ISR). He was also a co-founder of the NASA Center for Commercial Development of Space on Hybrid and Satellite Communications Networks established in 1991 at Maryland as an off-shoot of the ISR. He served as Co-Director of that Center from 1991 to 1994. He also held the endowed Cynthia Kim Eminent Chair of Information Technology and in 2012 he has been designated as Distinguished University Professor at the University of Maryland. Since 2021 he has become Professor Emeritus.

His interests are in the areas of communication theory, communication systems and networks, queuing systems, signal processing, and satellite communications, cross-layer design, and has initiated research in the field of Ad Hoc wireless networks.  Most recently he has spearheaded the area of semantic communications, including the Age of Information.

He has been the President of the Information Theory Society of the IEEE (1987) and has served on its Board of Governors almost continuously from 1981 until recently.  He was elected to the Board of Directors of the IEEE in 1989 and 1990. He won the IEEE Donald E. Fink Prize Paper Award (1992) and he was the first recipient of the Sigmobile Award of the ACM (Association of Computer Machinery) for contributions to wireless communications in 1997. He has been an IEEE Life Fellow. He has served on the Editorial Boards of the IEEE Transactions on Automatic Control, IEEE Transactions on Information theory, the Journal of Wireless Networks, and the International Journal of Satellite Communications. He has been the Editor-in-Chief of the Journal of Communications and Networks since 2016.

Europe has taken a clear stance that we want trustworthy AI.
This talk presents what trustworthy AI is as well as how to make
synthetic data generation and large-language models more trustworthy.

Fredrik Heintz is a Professor of Computer Science at
Linköping University, where he leads the Division of Artificial
Intelligence and Integrated Computer Systems (AIICS) and the Reasoning
and Learning lab (ReaL). His research focus is artificial intelligence
especially Trustworthy AI and the intersection between machine reasoning
and machine learning. Director of the Wallenberg AI and Transformative
Technologies Education Development Program (WASP-ED), Director of the
Graduate School of the Wallenberg AI, Autonomous Systems and Software
Program (WASP), Coordinator of the TAILOR ICT-48 network developing the
scientific foundations of Trustworthy AI, Coordinator of the TrustLLM
project, Vice President for AI Research Adra the AI, Data, and Robotics
partnership, and President of the Swedish AI Society. Fellow of the
Royal Swedish Academy of Engineering Sciences (IVA).

The field of computer vision is a sub area of AI, and it has its roots
in the modeling of the human visual system (HVS). It is commonly
accepted that about 80% of what we perceive is vision-based (Ripley and
Politzer 2010), but modeling vision is a systematically underestimated
scientific challenge – an implication of Moravec’s paradox, “we’re
least aware of what our minds do best” (Minsky 1986). The highly
intuitive nature of the HVS makes it difficult for us to understand the
myriad of interdisciplinary problems associated with computer vision.

The research at the Computer Vision Laboratory (CVL) has a strong focus
on theory and methods, in particular within machine learning, signal
processing, and applied mathematics. The resulting methods are applied
in fields where technical systems are supposed to coexist with and
therefore predict actions of humans. Self-driving cars sharing road
space and interacting with humans, sustainable forestry and
agriculture, monitoring of greenhouse gases as well as classification
and monitoring of animals are some application domains.

My talk will cover a wide range of topics within machine learning for
computer vision and robot perception: Few-shot and weakly supervised
learning, geometric deep learning, semi-supervised and incremental
learning, scene flow estimation, uncertainty representation, as well as
video and semantic segmentation.

Michael Felsberg received the PhD degree from Kiel University, Germany,
in 2002, and the docent degree from Linköping University, in 2005. He
is a full professor with Linköping University, Sweden. His research
interests include, besides visual object tracking, video object and
instance segmentation, classification, segmentation, and registration
of point clouds, as well as efficient machine learning techniques for
incremental, few-shot, and long-tailed settings.

The motivation for this project is the increasing requirements on audio
products for hearing aids. We aim for a set of optimized methods for
monitoring of recorded electroencephalogram measurements in order to
estimate and track sound processing in the brain. We attack the problem
of complex listening environments. The resulting algorithms will provide
a better understanding for how sound is processed at different stages in
the brain, opening up for future advanced hearing aid solutions. The
experiments and technical solutions are developed through the
collaboration with researchers at Eriksholm Research Centre, a part of
the hearing aid manufacturer Oticon A/S.

Maria Sandsten is professor in Mathematical Statistics. Her speciality
is optimization of time-frequency methods for spectral feature
extraction.  Current research work is on brain computer interfaces for
sound processing, machine learning for ecological monitoring and
acoustic vibrations from mining activities. She is also a project leader
in the Vinnova competence center Sentio – Integrated Sensors and
Adaptive Technology for Sustainable Products and Manufacturing.

In the rapidly evolving field of robotics, a significant shift is taking
place: the move from traditional industrial applications to dynamic
human-centered environments. This transition demands robotic systems
that are not just safe and collaborative but also highly adaptive and
capable of learning through interaction from their surroundings. In this
talk, I will focus on adaptive real-time solutions for reactive
trajectory generation, force-based robotic manipulation and grasping
implemented on collaborative robots but also learning-based solutions
for problems that involve deformable objects showcasing the complex
dynamics at play. I will discuss how haptic sensing can be utilized to
address uncertainty and facilitate adaptability, and how dual arm robots
can show increased flexibility and independence from a structured
environment. To conclude, I will highlight the essential synergy between
navigation, manipulation, and grasping, emphasizing the intricate
challenges involved in developing a cohesive robotic system that
seamlessly coordinates the functionalities of hand, arm, and body within
complex dynamic environments.

Yiannis Karayiannidis (Docent 2017, Phd Eng. 2009,  Diploma Eng. 2004)
is an Associate Professor (ELLIIT Recruited and WASP affiliated Faculty)
with the Department of Automatic Control, Faculty of Engineering, Lund
University since 2022. He was previously affiliated with KTH, Royal
Institute of Technology, from 2011 to 2020, and Chalmers University of
Technology, from 2015 to 2022. His research interests include robot
control, robotic manipulation in human-centered environments, dual arm
manipulation, force control, robotic assembly, control of physical
human-robot interaction, multi-agent robotic systems, adaptive control
and nonlinear control systems. He serves as Associate Editor for the
IEEE Robotics and Automation Letters, IEEE/RSJ International Conference
on Intelligent Robots (IROS) and Systems and the European Control
Conference (ECC), and has been a reviewer for the major scientific
journals specialized in robotics and automatic control. He also serves
as the treasurer of the IEEE Robotics Chapter in Sweden. He was the PI
for national projects such as DARMA and DARMA_bridge by WASP, CHROMA by
VR, but also for the H2020 SARAFun project representing KTH.

Graphs can be used to represent data with complex structure and
dependencies in a wide range of application domains, ranging from
chemistry to communication networks and spatio-temporal prediction.
Graph neural networks (GNNs) have proven to be a powerful model class
for various prediction tasks involving such data. In this talk I will
give a brief overview of some of the ongoing projects involving GNNs
within the ELLIIT environment at the Statistics and Machine Learning
Division at Linköping University, including: (i) speeding up phase
diagram generation for materials discovery, (ii) constructing flexible
priors for graph data by bridging between GNNs and probabilistic
graphical models, and (iii) using hierarchical GNNs for weather
forecasting.

Based on work done by ELLIIT PhD students Filip Ekström Kelvinius and
Joel Oskarsson.

Fredrik Lindsten is a Senior Associate Professor in Machine Learning at
the Division of Statistics and Machine Learning, Department of Computer
and Information Science, Linköping University, Sweden. He is an ELLIIT
Recruited Faculty and Faculty of WASP. He received his MSc degree in
Applied Physics and Electrical Engineering in 2008 and a PhD in
Automatic Control in 2013, both from Linköping University. In 2014 and
2015 he was a Postdoctoral Research Associate at the Department of
Engineering, the University of Cambridge, UK. During spring 2012 he was
a Visiting Student Researcher at the Statistical Artificial Intelligence
Lab at the University of California, Berkeley, USA and during spring
2015 he was a Visiting Scholar at the Department of Statistics, the
University of Oxford, UK. He has received the Ingvar Carlsson Award by
the Swedish Foundation for Strategic Research, and the Benzelius Award
by the Royal Society of Sciences in Uppsala. Lindsten’s main research
interests are in statistical machine learning and computational statistics.

– The cloud journey: How it all came together, where we are now, and
where it might lead us

– The importance of automation: Everything everywhere needs automation-,
needs to handled as code (x-OPS), and needs to be declarative.

– How will the networks provide differentiated network experience, and
will network programability provide such experience.

– How could the IT-, OT-, and Communication industries become better
aligned around technologies, standards- and best practices.

Jonas has 35+ year’s experience with communication technologies- and architectures in various contexts. 

Many years ago, Jonas lead the architecture development for the WCDMA application infra-structure, since then Jonas has been continuously engaged in pushing the evolution of the communication industry’s technologies-, architectures-, and capabilities including roles and tasks such:

– At the Millenium, Jonas Co-founded a joint venture between Ericsson and Juniper Networks, where he served

  as CTO, as well as Vice President of product management and engineering, leading the development of the

  first ever high performance router-based Mobile packet Gateway (GGSN).

– In 2003, Jonas lead the architecture definition- and development of Ericsson’s first all Ethernet based communication infrastructure, base for all of Ericsson’s core network applications. This platform shared many of the properties that you would see in today’s private cloud solutions.

– In 2009, Jonas co-founded “Ericsson In-game communication”, a wireless communication platform for the on-line gaming industry, the platform was launched at the Mobile World Congress 2010, and later came to be part of several on-line gaming platforms before it was spun out from Ericsson. 

Since 2013, Jonas have had several leading roles in Ericsson’s journey towards an all cloud native paradigm, including both the infra-structure and the applications, as well as Ericsson’s engagement in Open-source in this area.

Jonas today holds the position as Chief architect for the “Ericsson cloud infrastructure”, in this position he oversees the architecture and the technology strategies for Ericsson’s virtualization- and container platforms.

The operation of buildings accounts for 30% of final global energy
consumption, and 40% thereof is used for Heating, Ventilation and Air
Conditioning (HVAC). Optimizing HVAC design and operation to reduce
energy consumption and carbon emission footprint in buildings is
therefore one of our global grand challenges to reach sustainability
goals. Carrier, an HVAC market leader, is in the middle this on-going
transition to more sustainable and efficient energy systems for
buildings. Design and energy efficiency improvements to equipment units
(chillers, heat pumps etc.) remains important and consistently result in
single digit improvements. However, double digit energy savings are
feasible by optimizing energy efficiency of building energy systems,
where sources, sinks and storage of energy is coordinated as to minimize
energy, carbon and/or financial objectives. To realize those savings,
new methods and tools are needed in design, control and operation of
such systems. In this presentation, challenges in building energy
management will be reviewed and opportunities to apply advanced control,
system optimization and machine learning will be highlighted.

Johan Åkesson serves as Associate Director for Computational Engineering
within the Systems, Controls and ML/AI Center of Excellence at Carrier
Central Engineering. In this role, he leads a team of engineers engaging
across all branches of Carrier worldwide to deploy optimization-based
methodology in product design, decision support tools, advanced
controls, and system operation. He previously held positions as
Assistant Professor at the department of Automatic Control, Lund
university, and Chief Technology Officer at Modelon, both in Sweden. His
current location is Lund, Sweden. His areas of expertise in the past 15
years include model-based engineering, optimization of physical systems
for design and controls, numerical optimization algorithms, software
development of modeling and optimization tools, and cloud deployment. He
is a long-standing member of the Modelica Association and has actively
contributed to the development of the Modelica and FMI standards. In his
former position within Modelon he was part of establishing the software
development branch of Modelon with responsibilities for product
development and delivery of Modelon products. Åkesson has a Master’s
degree in Computer Science from Lund university, Sweden, and a PhD
degree in Automatic Control also from Lund university.

We describe applications of symbolic computation towards automating the formal analysis of while-programs implementing polynomial arithmetic.

We combine methods from static analysis, computer algebra and probability theory in order to automatically derive loop invariants and capture how the value distributions of loop variables influence the functional behavior of loops. Our results yield computer-aided solutions in support of formal software verification, compiler optimization, and probabilistic programming. We demonstrate the practical use of our techniques via illustrative example and using our Polar static analyzer. 

Laura Kovacs is a full professor of computer science at the TU Wien, leading the automated program reasoning (APRe) group of the Formal Methods in Systems Engineering division. Her research focuses on the design and development of new theories, technologies, and tools for program analysis, with a particular focus on automated assertion generation, symbolic summation, computer algebra, and automated theorem proving. She is the co-developer of the Vampire theorem prover and a Wallenberg Academy Fellow of Sweden. Her research has also been awarded with a ERC Starting Grant 2014, an ERC Proof of Concept Grant 2018, an ERC Consolidator Grant 2020, and an Amazon Research Award 2020. Recently, she received financial support from LEA Frauenfonds to disseminate unplugged computer science to elementary schools, while organising computer science workshops with school children at the TU Wien.

– Håkan Grahn: GPAI – General Purpose AI Computing
– Elina Rönnberg: Decomposition to Tackle Large-scale Discrete
Optimisation Problems
– Emma Söderberg: Gazing at Code Reviews
– Simin Nadjm-Tehrani: XG communication: Security Issues and Solutions
in Store
– Oscar Gustafsson: Accelerating robust AI (and DSP) for space (and
other critical scenarios)

– Ehsan Miandji: FROST-BRDF: A Fast and Robust Optimal Sampling
Technique for BRDF Acquisition
– Christian Kowalkowski: B2B Data Sharing for Industry 4.0 Machine Learning
– Carmela Bernardo: Analyzing leadership strategies in climate change
negotiations with a social power game
– Johan Lundgren: Real-Time Near-Field Measurements of mmWave Devices
Using a Metasurface and IR Camera
– Håkan Johansson: Low-Complexity Memoryless Linearizer for
Analog-to-Digital Interfaces

– Viktor Larsson: Multi-view Matching and Reconstruction of Line Segments
– Krzysztof Bartoszek: Fast inference for hierarchically dependent data
– Simon Kristoffersson: Making the Flow Glow – Robot Perception under
Severe Lighting Conditions using Normalizing Flow Gradients
– Tom Ziemke: Human interaction with Social Robots & Automated Vehicles
– Amir Aminifar: Energy-Efficient Distributed Inference for Internet of
Things (IoT) Systems