Lars Bramsløw

Ph.D., Principal Scientist

Eriksholm Research Centre, Oticon A/S

Denmark

Modern hearing aids pack incredible signal processing and wireless features on a single-mW power budget. They must be designed to be worn all day, providing tangible end user benefits for different levels of hearing loss and interface to dedicated smartphone apps. Each hearing loss is special, and the signal processing incl. hearing loss compensation must be sophisticated and individually tailored to address the user challenges, especially communicating in noisy environments: ‘the cocktail party problem’.
Current hearing aid development is based on deep insight into hearing and hearing loss and a broad knowledge of signal processing, within dynamic range compression, noise reduction, beamforming, anti-feedback, artificial intelligence etc. All this must be integrated to a stable and pleasant audio processing system delivering optimum performance in all possible listening scenarios. The development follows a patient, iterative and pragmatic cycle. For listening evaluation in this cycle, a combination of signal metrics, informal listening and formal listening tests is applied. Traditionally, we focus on speech intelligibility in noise and sound quality, but newer outcome measures are also used, such as listening effort, attention, and physiological measurements.
Despite the hardware restrictions, the main limitation to better treatment of hearing loss todayis fundamental research on hearing, hearing loss and how it relates to the daily life of the user. Hence, more hearing research is needed to improve devices even more.

Lars Bramsløw, PhD, is a Principal Scientist at the Eriksholm Research Centre, part of Oticon. He holds an M.Sc. and a Ph.D. degree, both from Technical University of Denmark, in 1986 and 1993. The Ph.D. was carried out at Eriksholm on the topic of sound quality and neural networks.
Lars has 30 years of extensive experience in acoustics, hearing science and hearing aid research and development, including employment at House Ear Institute in Los Angeles, Eriksholm and Oticon headquarters in Smørum, Denmark. He is currently working on improved fitting of hearing aids for the individual and the application of deep learning algorithms in hearing health care. He is also the current president of the Danish Acoustical Society.

Giuseppe Tortora

Founder and CEO of ABzero

ABzero

Italy

Applications of Artificial Intelligence algorithms can be useful for real life medical scenarios. In this tutorial, drone delivery applications driven by AI are presented as case studies at hospital facilities. The experience of an innovative solution of drone delivery originated as a technology transferred from the Smart Medical Theatre Lab to an innovative Italian startup company, ABzero will be reported. Research and innovation came together in joint collaborations between universities and companies, as was the case with University of Poznan, in which Artificial vision algorithms have been improved for being applied to the innovative delivery system in order to improve it. With an application-oriented and entrepreneurial approaches, this workshop will introduce the concepts of drone delivery of medical materials, such as blood, blood components and organs.

Giuseppe Tortora, founder and CEO of ABzero, an Italian startup company. After a university and research training at the Scuola Superiore Sant'Anna, Pisa, Imperial College London, Carnegie Mellon University Pittsburgh, in the fields of medical robotics, minimally invasive robotic surgery, microrobotics, Giuseppe has gained entrepreneurial experience since 2014. He was awarded with many international awards for entrepreneurs. Giuseppe Tortora had experience in a previous medical startup.

Ass. Prof. Christos P. Antonopoulos

Electrical & Computer Engineering Department

University of Peloponnese

Greece

Cyber-physical Systems of Systems (CPSoS) are large complex systems where physical elements interact with many distributed computing elements and human users. The main challenges associated with the design operation of reliable connected CPSs stream from continuously increasingly demands for resources availability, novel service of high added value and product quality, while adhering to requirements for low cost, low power and competitiveness in the global market. This paper puts forward a set of CPSoS systems aiming at developing models, architectures and software tools for allocating computing resources to CPS end devices, while autonomously detecting what cyber-physical processes will be performed by the device components including heterogeneous CPU elements and communication interfaces, GPUs, FPGA fabrics, and Software stacks. Collaborating in a decentralized way and ensuring the exchange of tasks and data without central intervention, the proposed projects are leveraging state of the art artificial intelligence technologies to improve their performance, reliability, robustness as well as security.

Ass. Prof. Christos P. Antonopoulos received his Diploma and PhD at Electrical Engineering and Computer Technology from ECE Department of Electrical Engineering and Computer Technology, University of Patras Greece in 2002 and 2008 respectively. From 2019 he holds the position of Assistant Professor at the ECE Department of University of the Peloponnese, Greece. From 2002, he has been involved in more than 15 European projects (FP5,6,7 and Horizon 2020) and more than 6 Greek National research projects holding key positions both technical as well as managerial while he has participated in preparation of a high number of research project proposals.He has published high number (>100) of research papers and 13 book chapters in international journals and conferences which have received over 900 citations while he has served as editor to two Springer books. From 2005 he has worked and collaborated as Senior Researcher for several research institutions (Industrial System Institute (ISI) Patras, ECE Dept University of Patras, Technological and Educational Institute of Western Greece, Hellenic Open University) and industrial companies (INTRACOM-Telecom, NOESIS). Over the years, he has reviewed a high number of submitted paper for high quality IEEE, ACM, Elsevier, Springer and Hindawi, MDPI Journals and conferences and he has served as organizer, PC/TC member and session chair in several prestigious international conferences. Additionally, from 2003 he has worked continuously as adjunct assistant professor for various departments at Greek Universities and Technological Institutes.
His main research interests include: Wireless Networks, Network Communication Protocols, Network Simulation, Performance Evaluation, Embedded Software, Sensor Networks, Wireless Broadband Access Networks, Embedded System Architecture and Programming, Cross-Layer protocols, Wireless Networks power optimization, Internet of things, Cyberphysical systems.

Miha Gjura

Technical Specialist

Red Pitaya

Slovenia

Join us for an engaging tutorial highlighting Red Pitaya as the perfect tool to accompany students on their signal processing education and research journey. Designed for signal processing academics, researchers, and educators, this session will showcase the user-friendly and practical applications of Red Pitaya, enabling students to develop expertise in signal processing.
Discover how easy it is for students to get started with Red Pitaya, as we walk you through the seamless connection via a browser and SSH. Witness the platform's capabilities in action as students make their first signal acquisitions using the Red Pitaya Oscilloscope and signal generator.
The ability to perform signal acquisition and generation through SCPI commands empowers students to set up custom signal acquisition with confidence, providing a solid foundation for their signal processing exploration.
As students progress in their journey, they can unleash the full potential of Red Pitaya by programming its FPGA using Vivado. This opens up exciting opportunities for advanced signal processing projects and research endeavors.
Get inspired with real-world examples of projects that highlight Red Pitaya's adaptability, from radars to other innovative applications. Red Pitaya serves as the ideal companion for students as they undertake research projects, advancing their signal processing expertise.
For educators, Red Pitaya offers a powerful teaching tool to engage students at every level of their signal processing learning. For researchers, Red Pitaya provides a versatile and capable platform to explore cutting-edge signal processing applications.
With informative demos and an interactive Q&A, this online presentation is a must for educators seeking powerful teaching tools and researchers looking for a reliable companion in their signal processing investigations. Embrace the future of signal processing education and research with Red Pitaya, empowering students and researchers to excel in this dynamic field.

Miha Gjura is a Technical Specialist at Red Pitaya, a Slovenian company renowned for developing versatile open-source software-defined instruments and FPGA development platforms. His expertise lies in OS testing, Python programming, FPGA, and C development, enabling significant contributions to enhance and optimize Red Pitaya's innovative solutions. Miha's dedication to education is evident in preparing workshops for students using the Red Pitaya platform, empowering the next generation of engineers and scientists with practical insights. He recently led a hackathon challenge at UCSD, showcasing the innovative capabilities of Red Pitaya's products and fostering exploration of new technological frontiers. Furthermore, Miha serves as a valuable resource for Red Pitaya users, offering troubleshooting assistance and support to maximize the potential of their devices in projects and research endeavours.

Prof. Thomas M. Coughlin

President, Coughlin Associates

United States

Digital storage and memory are critical elements in modern data processing, enabling training of large AI models, scientific and engineering data analysis, smart factories and most consumer applications. This talk will discuss available storage and memory technologies (DRAM, SRAM, NAND Flash, SSDs, HDDs, Magnetic Tape and Optical Recording as well as emerging non-volatile memory technologies), including how they are used and projections for their future of digital storage and memory.

Tom Coughlin, President, Coughlin Associates is a digital storage analyst and business/ technology consultant. He has over 40 years in the data storage industry with engineering and senior management positions. Coughlin Associates consults, publishes books and market and technology reports and puts on digital storage-oriented events. He is a regular contributor for forbes.com and M&E organization websites. He is an IEEE Fellow, 2023 IEEE President Elect, Past-President IEEE-USA, Past Director IEEE Region 6 and Past Chair Santa Clara Valley IEEE Section, and is also active with SNIA and SMPTE. For more information on Tom Coughlin go to www.tomcoughlin.com.

Prof. Patrick Dewilde

Institute for Advanced Study

Technische Universität München

Germany

Orthogonal filtering consists in using orthogonal or unitary transformations to process data, in order to extract interesting characteristics or to achieve desirable control tasks. It is the basic step in many well-conditioned numerical operations on data, in particular quadratic optimal control, Kalman filtering, smoothing, spectral factorization, selective filtering, feature extraction, back propagation etc. The simplest numerical realization of an orthogonal filter is known as `QR-factorization’, while in classical filtering theory it is known as `minimal phase extraction’. Mathematicians often use the term `inner-outer’ and `outer-inner’ factorization, dating back to Hardy-space theory in the early 20th century. The technique is very generally applicable, even in time-variant and non-linear situations (although the latter has not been fully explored so far.) Conversely, it turns out that many classical problems that have been treated with `brute force methods’ can very elegantly and even simply be solved just using orthogonal filtering directly. The tutorial will show how that works.

Patrick Dewilde (EE `66 KULeuven, Belgian Lic. Math. `68 and PhD `70 Stanford University) has been a professor in electrical engineering at the Technical University of Delft for 31 years, director of the Delft Institute for Micro-electronics DIMES for ten years, chairman of the Technology Foundation STW (a major Dutch research funding agency) for eight years and director of the Institute of Advanced Study of the TU Munich for five years. His research, published in the international scientific literature and some books, has focused on mathematical issues related to the design, control and operations of dynamical systems in general and in particular circuits and systems for signal processing. He is an IEEE Fellow since 1981, an elected member of the Dutch Royal Academy of Arts and Science, has been elevated to the rank of Knight of the Dutch Lion and is presently a honorary professor both at the Technische Universität München and the Technical University of Wroclaw.

Prof. Dr. Marcel Rupf

Institute of Signal Processing and Wireless Communications (ISC)

ZHAW School of Engineering

Switzerland

Today, FMCW-MIMO-based (mm-wave) radar sensors are widely used in automotive and industry because the FMCW-technique has some important advantages over the pulse-echo principle. In this tutorial lecture, it will be shown, that the range, the doppler speed and the angle-of-arrival of multiple targets can be determined efficiently by using FFTs. Additionally, some more challenging topics like clustering, tracking and classification of targets will be addressed based on classical DSP-algorithms and Deep Neural Networks (DNN). Finally, some radar sensing applications will be shown, e.g. for bird migration monitoring, vehicle tracking and gesture recognition.

Prof. Dr. Marcel Rupf
Academic Education
1987 Diploma in Electrical Engineering from ETH Zürich, Dipl. El. Ing. ETH
1993 PhD in Technical Science at ETH Zürich, Advisor: J.L. Massey
1993-1995 Postdoc at the IBM Research Lab in Rüschlikon, Switzerland
Work Experience in 3 Companies (Region of Zurich)
1995-2003 Consultant, Project Manager and Head of R&D in Mobile Communications
Zurich University of Applied Science (ZHAW)
since 2003 Lecturer for Digital Signal Processing and Wireless Communications
2007 Establishing Center for Signal Processing and Wireless Communications ZSN
2007-2017 Head of ZSN
2017 Establishing Institute of Signal Processing and Wireless Communications ISC (http://www.zhaw.ch/isc/) out of ZSN
since 2017 Head of ISC

Prof. Bożena Kostek

Audio and Acoustics Laboratory

Faculty of Electronics, Telecommunications, and Informatics

Gdansk University of Technology

Poland

This tutorial lecture is dedicated to addressing the still not fully achieved potential in automatic music and speech processing. There are parallels between music and speech as both are audio signals; however, they differ much in detail. One of such parallels are automatic music transcription and text-to-speech technology; both are very advanced in research and technology but suffer from challenges related to polyphony and differentiated articulation in music, whereas in speech: differentiated accents, L2 speakers’ pronunciations, diarization, just to name a few, cause problems. Another challenge concerns affective computation and predicting emotions in music and speech with sufficient accuracy. These challenges lie in the notion that we would like to get better results with machine learning-based classification than human expertise could deliver.

In this lecture, first, bases of automatic music and speech processing are to be shortly reviewed. Then, the state-of-the-art in machine learning, including both baseline and deep learning methods, is briefly examined in the context of music and speech processing. Also, advances in available technology related to such issues are presented. Finally, examples of research performed in the discussed areas are shown.

Bożena Kostek is a professor and head of the Audio and Acoustics Laboratory at Gdansk University of Technology (GUT), Poland. She is a corresponding member of the Polish Academy of Sciences, a Fellow of the Acoustical Society of America, and the Audio Engineering Society. She published more than 600 scientific papers in journals. She led many research projects. Prof. Kostek acted as Editor-In-Chief for the Archives of Acoustics (JCR journal), the Journal of the Audio Engineering Society (JAES), and in 2013 she was appointed as Associate Editor of the IEEE Transaction of Audio, Speech and Language. Under her guidance, 18 Ph.D. students supported their doctoral theses and 280 M.Sc. and Eng. works. She is the recipient of many prestigious awards, including two 1st prizes of the Prime Minister of Poland, several prizes of the Minister of Science, and the award of the Polish Academy of Sciences.

Prof. Andrzej Materka

Institute of Electronics

Lodz University of Technology

Poland

Vascularity diseases comprise one of the major health problems worldwide. Their diagnosis, therapy planning, treatment, and etiology understanding require personalized geometrical modeling of the vessel trees and measurement of their parameters, e.g. local lumen radius, volume of calcification or centerline course. Arteries’and veins’ wall surface models are essential for visualization and for blood flow numerical simulation – in support ofbiomedical research and education.Medical imaging is the main technique that makes those measurements possible in a least invasive way, with 3D magnetic resonance (MR) and computed tomography (CT) modalities being the most popular. This tutorial is focused on methods of automated processing of 3D images for objective quantification of blood vessel lumen. One of the aims of the search for such methods is to release medical experts from the tedious, time-consuming (and overall subjective) annotation of the enormous number of voxels in 3D space. Other expectations include better repeatability, objectivity and accuracy/precision of the obtained diagnostic data and shorter time of their extraction from images. Accurate segmentation of the vascular objects in the image is a challenging task – the images contain noise, artefacts, the vessels feature high shape complexity and are closely surrounded by other tissues and organs of similar appearance. The vessel branches’ diameter take values in a wide range – from tens of millimeters to tens of micrometers – while the resolution of images is limited and can not be improved without increasing the noise level or time of examination. The noise introduces uncertainty to segmentation results, expanded acquisition time may result in significant artefacts caused by movement of organs and patient body. A widely adopted solution – acquiring images composed of anisotropic voxels (small in-plane elements extending deep into a few times thicker slice) – is a compromise which does not help in resolving details of 3D shapes. Superresolution preprocessing might be helpful in reducing the segmentation errors in such cases. There are two general approaches to vessel segmentation – via 2D cross-sections perpendicular to the vessel centerline and by direct 3D volumetric segmentation. The first one is usually preceded by image “vesselness” filtering to define the centerline, the second may use long-lasting iterative computations, e.g. those of the level-set algorithm. There is an open-source and free software available for 3D vascular image segmentation, e.g. SimVascular or ITK-Snap, as well as a few free-access databases of annotated vasculature images. Reference to those resources and examples of their use will be provided. Centerline-based methods and code, developed at TUL Institute of Electronics in collaboration with researchers from Jena University and Medical University of Łódź will also be illustrated. Various schemes of 2D cross-sections segmentation and quantification will be compared, with application to computer-simulated 3D images, MR images of 3D-printed vascularity models, human coronary arteries visualized in X-ray CT volumes, and 3D MR images of human brain blood vessels.

Andrzej Materka received the M.Sc. degree from Warsaw University of Technology, the Ph.D. degree from Łódź University of Technology (TUL), and the Dr.Hab (D.Sc.) degree from Wrocław University of Technology, respectively in 1972, 1979, and 1985. In 1996 he was awarded the title of professor in technical sciencesby the President of Poland.
In 1972-74 he was an engineer at R&TV Broadcasting Stations, Łódź, and since 1974 he has been employed at TUL Institute of Electronics, serving as its director in 1995-2015. In 1980-82 he was a Monbusho scholar at Shizuoka University, Hamamatsu, Japan, and in 1992-94 worked as a senior lecturer at Monash University, Faculty of Electrical and Computer Systems Engineering, Melbourne, Australia. In 2002-8 he was a dean of TUL Faculty of Electrical and Electronic Engineering. His research interests include numerical modeling of semiconductor devices, microwave techniques, medical electronics, digital signal processing and image analysis, artificial neural networks, brain-computer interfaces and geometric reconstruction of bloodvessels from images. Professor Materka published over 200 technical articles and 6 monographs, cited over 1200 times (h=18). He is a Life Senior Member of the IEEE.

Prof. Sanja Lazarova-Molnar

Institute of Applied Informatics and Formal Description Methods

Karlsruhe Institute of Technology, Karlsruhe

Germany

Simulation models are often developed manually with significant dependence on expert knowledge. Nowadays, however, availability and prevalence of data has the potential to completely transform these traditional simulation modeling processes. The advances in data and process mining combined with data modeling techniques promise much in this direction. What can be automated and what has to be in the hands of experts in the data-driven simulation modeling? The tutorial will focus on these questions and provide directions and answers to addressing them.

Sanja Lazarova-Molnar is a Professor at the Institute of Applied Informatics and Formal Description Methods, Karlsruhe Institute of Technology. She is also a Professor at the University of Southern Denmark, where she leads the research group Modelling, Simulation and Data Analytics. She is a Senior Member of The Institute of Electrical and Electronics Engineers (IEEE), and currently serving as Director-at-Large on the Board of Directors of The Society for Modeling & Simulation International (SCS). Furthermore, she is Chair of IEEE Denmark and Vice-Chair of IEEE Denmark Women in Engineering Affinity Group. As of latest, her research focus is on data-driven simulation in various contexts and for the various simulation paradigms, and reliability modeling of cyber-physical systems, such as manufacturing systems and smart buildings. Her email address is sanja.lazarova-molnar@kit.edu.

Prof. Gilbert Strang

Department of Mathematics

Massachusetts Institute of Technology

United States

The equation Ax = 0 tells us that x is perpendicular to every row of A – and therefore to the whole row space of A. This is fundamental, but singular vectors v in the row space do more. (1) the v’s are orthogonal (2) the vectors u = Av are also orthogonal (in the column space of A). Those v’s and u’s are columns in the singular value decomposition AV = UΣ. They are eigenvectors of A^TA and AA^T , perfect for applications.
We can list 10 reasons why orthogonal matrices like U and V are best for computation — and also for understanding. Fortunately the product of orthogonal matrices V₁ V₂ is also an orthogonal matrix. As long as our measure of length is ||v||² = v₁² + ... + v_n², orthogonal vectors and orthogonal matrices will win.

Gilbert Strang was an undergraduate at MIT and a Rhodes Scholar at Balliol College, Oxford. His Ph.D. was from UCLA and since then he has taught at MIT. He is a member of the National Academy of Sciences. His classes are the most watched on MIT’s opencourseware ocw.mit.edu and his books include Introduction to Linear Algebra (5th edition) and the 2019 book Linear Algebra and Learning from Data : math.mit.edu/learningfromdata. The new book for 2020 is Linear Algebra for Everyone.

Prof. Dr.-Ing. habil. Michael Hübner

Computer Engineering Group

Brandenburg University of Technology - Cottbus - Senftenberg

Germany

Cyberphysical systems need to handle complex applications in different domains. The challenge is the parameterization and compositions of resources of such systems which needs to be done traditionally at design time. This requires to explore a large design space and is therefore not leading to an optimized setup. Run-time adaptive systems can find a optimal point of operation at run-time which is an advantage. This talk shows possible architecture which allow such a flexibility and discusses future solutions.

Prof. Dr.-Ing. habil. Michael Hübner (male) (IEEE Senior Member) is full professor at the Brandenburg University of Technology - Cottbus - Senftenberg. He is leading the Computer Engineering Group since october 2018. From 2012-2018 he led the Chair for Embedded Systems for Information Technology (ESIT) at the Ruhr University of Bochum (RUB). He received his diploma degree in electrical engineering and information technology in 2003 and his PhD degree in 2007 from the University of Karlsruhe (TH). Prof. Hübner did his habilitation in 2011 at the Karlsruhe Institute of Technology (KIT) in the domain of reconfigurable computing systems. His research interests are in reliable and dependable reconfigurable computing and particularly new technologies for adaptive FPGA run-time reconfiguration and on-chip network structures with application in automotive systems, incl. the integration into high-level design and programming environments. Prof. Hübner is main and co author of over 200 international publication. He is in the steering committee of the IEEE Computer Society Annual Symposium on VLSI, organized more than 25 events like workshops and symposia, and is active as guest editor in many journals like e.g. IEEE TECS, IEEE VCAL and IEEE TNANO.

Prof. Georgios Keramidas

Aristotle University of Thessaloniki

Embedded Systems Design and Applications Laboratory

University of Peloponnese

Greece

SMART4ALL is a four-year Innovation Action project funded under Horizon 2020 framework under call DT-ICT-01-2019: Smart Anything Everywhere – Area 2: Customized low energy computing powering CPS and the IoT. The target of the project is to establish a unique Pan European network of Digital Innovation Hubs that will not only support innovation and reveal business opportunities across South, Eastern and Central Europe, but will build capacity via the development of self-sustained, cross-border pathfinder application experiments. The project will provide a total funding of 2,2 Mio Euros via 9 open calls and it will support 88 cross-border pathfinder application experiments from European consortia. Each experiment will get funding up to €80,000 and it will be supported with novel coaching services from world lead experts in ethics, technology, funding and business development. Apart from this, SMART4ALL sets forward a unique concept called Marketplace-as-a-Service (MaaS). Maas is a one-stop-smart-shop for startups, SMEs and slightly bigger companies. This presentation will introduce SMART4ALL and it will concentrate on the “Prepare for Growth” services offered by Maas.

Georgios Keramidas is an assistant professor at the Aristotle University of Thessaloniki and a senior researcher at the Embedded Systems Design and Applications Laboratory at the University of Peloponnese, Greece. Dr. Keramidas serves as the technical coordinator of the SMART4ALL project. His main research interests are in the area of low-power processor/memory design, multicore systems, VLIW/multi-threaded architectures, network and graphic processors, reconfigurable systems, power modelling methodologies, FPGA prototyping, and compiler optimizations techniques. He has published more than 70 papers, two books, and he also holds 10 US patents (4 more patents are under evaluation). His work received more than 1000 citations. During the last years, Dr Keramidas has participated in nine research projects funded by European Commission and in five national projects either as project coordinator, technical coordinator, work packager leader, or as senior researcher. Dr. Keramidas is a member of ACM, a regular reviewer and program committee member in high-quality conferences, workshops and transactions and a member of the HiPEAC European Network of Excellence.

Mr Antonio Montalvo

Telecommunications Engineer & MBA

Senior Project Manager at FundingBox Accelerator

Poland

SMART4ALL is a four-year Innovation Action project funded under Horizon 2020 framework under call DT-ICT-01-2019: Smart Anything Everywhere – Area 2: Customized low energy computing powering CPS and the IoT. The target of the project is to establish a unique Pan European network of Digital Innovation Hubs that will not only support innovation and reveal business opportunities across South, Eastern and Central Europe, but will build capacity via the development of self-sustained, cross-border pathfinder application experiments. The project will provide a total funding of 2,2 Mio Euros via 9 open calls and it will support 88 cross-border pathfinder application experiments from European consortia. Each experiment will get funding up to €80,000 and it will be supported with novel coaching services from world lead experts in ethics, technology, funding and business development. This presentation will describe in detail the Open Calls, including their value proposition, the processes for eligibility (who can apply), evaluation and selection, the financial support and how to submit an application.

Antonio M. Montalvo (male) Project Manager at FundingBox. He holds a Master Degree in Telecommunications Engineering from the UPM (Universidad Politécnica de Madrid) and a Master in Business Administration from the Universidad PontificiaComillas. He started his career in the telecommunications sector, working from technical to managerial positions, in Europe and the United States. He participated in the EU-funded project which developed the High-Definition TV system for Europe. Later he moved to Latin America, where he was involved in the management of private and public projects for telecom, transport and housing infrastructure in countries such as Mexico, the Dominican Republic and Costa Rica. Back to Europe, during the last years, he has worked as Project Manager for different EU-funded projects, such as H2020-732947 frontierCities2, a FIWARE accelerator where he coordinated the grant management and financial support to third parties. He is currently managing three H2020 projects, dealing with the Open Call management: H2020-824964 DIH2, about the implementation of robotics for the agile production in manufacturing companies; H2020-76742 L4MS with the same aim as DIH2 but specifically devoted to logistics companies; and H2020-872614 SMART4ALL, about the technology transfer of any smart domain between countries of South, East and Central Europe.

Prof. Nikolaos S. Voros

Embedded Systems Design and Application Laboratory

Department of Electrical and Computer Engineering

University of the Peloponnese

Greece

Ambient Assisted Living (AAL) environments are rapidly shifting towards ICT based solutions. In this context, a wide range of consumer electronics technologies come into play ranging from robotics, embedded systems, sensors and communication infrastructures. Open access infrastructures and services can play a key role in the wide adoption of the next generation AAL coaching systems. In that respect, we present as service-oriented, easily expandable design paradigm that emphasizes on how heterogenous COTS ICT technologies can be used as enablers of existing and new AAL services. The services focus on Activities of Daily Life (ADL) monitoring algorithms, on facilitating the indoor everyday life activities of the end user emphasizing on energy multifaceted conservation and security provision.

Prof. Nikolaos S. Voros received his Diploma in Computer and Informatics Engineering, in 1996, and his PhD degree, in 2001, from University of Patras, Greece. His research interests fall in the area of embedded and cyberphysical system design. Prof. Voros is Associate Professor at Technological Institute of Western Greece, Department of Computer and Informatics Engineering where he is leading the Embedded Systems Design and Application Laboratory, which is a validated DIH of European Commission. During the last years, Prof. Voros has participated in more than 20 research projects funded by European Commission either as project manager or as scientific advisor. He has served as scientific coordinator of FP7 STREP Project 287733 ALMA and Horizon 2020 Project ARGO, and as ICT coordinator for FP7 STREP Project 287720 ARMOR and Horizon 2020 Project RADIO. Prof. Voros has published significant number of books and referred articles in international journals and conferences, while he is also editor of the books «System Level Design with Reuse of System IP», «System«System Level Design of Reconfigurable Systems-on-Chip», «Components and Services for IoT platforms: paving the way for IoT standards», «Advances in Aeronautical Informatics: Technologies Towards Flight 4.0» and «RADIO - Robots in Assisted Living: Unobtrusive, Efficient, Reliable and Modular Solutions for Independent Ageing» published by Springer. He was the General Chair of the IEEE Computer Society Annual Symposium on VLSI, which took place in Kefalonia, Greece on July 2010, while he also served as Program Co-Chair for 24th International Conference on Field Programmable Logic and Applications (FPL 2014), which took place at Technical University of Munich on September 2nd – 4th, 2014. On May 2018, he organized the 14th International Symposium on Applied Reconfigurable Computing which took place in Santorini, Greece.

Prof. Dr.-Ing. Heinrich Theodor Vierhaus

Fachgebiet Technische Informatik

Brandenburg University of Technology Cottbus-Senftenberg

Germany

Integrated circuit technology has fought against two specific “walls” during the last decade. First it was the “power wall”, followed by the “reliability wall”. Unfortunately, these walls are not independent, resulting in the need for on-line fault detection and correction at minimum cost in power. Duplication and even more triplication of hardware, in other words double-and triple modular redundancy, are heavily used today, since they are robust in terms of fault duration and timing in general, and they work without costly and complex re-design in modules such as embedded processor cores. Alternative methods have been developed, which detect faults by observing signal transitions by the wrong time, specifically for delay faults in logic and radiation-induced faults in flip-flops and registers. There are ways of combining these approaches with a promise for fault detection and correction at minimum cost in power. The tutorial introduces such techniques and shows their limits. Also combinations with more robust techniques are possible for so-called “self-dual” circuits. Finally the aspect of system level error resilience is addressed.

Heinrich Theodor Vierhaus received a diploma in electrical engineering from Ruhr-University Bochum (Germany) in 1975.
Lecturer for RF technology and electronic circuits with Dar-es-Salaam Technical College in Tanzania (East Africa) from 1975 to 1977.
Dr.-Ing. in EE from University of Siegen in 1983.
Senior researcher with GMD, the German National Research Institute for Information Technology from 1983 to 1996.
Professor for “Technische Informatik” (computer engineering) at Brandenburg University of Technology Cottbus since 1996.
He has authored or co-authored about 250 papers in the area of computer engineering, mainly related to IC test, technology, dependability and fault correction. He also contributed to three books in the area as an editor and an author. Since 2009, he has initiated and coordinated several projects on advanced education of doctoral students in the area of dependable systems, based on an international network of universities and research institutes.

Prof. Dr.-Ing. Michael Hübner

Fachgebiet Technische Informatik

Brandenburg University of Technology Cottbus-Senftenberg

Germany

The talk presents current and novel research about adaptive computing systems. These systems are able to adapt to the requirements of algorithms or quality of service requests from the user during run time. Additionally, next generation of embedded systems need to provide a high robustness e.g. by correcting faults which can occur on chip level. This can be achieved by a increased resilience.

Prof. Dr.-Ing. Michael Hübner has been working for over a decade in several domains, e.g. reconfigurable computing or embedded systems for Internet of Things. He is main- and co-author of over 300 scientific publications in international conference proceedings and journals. He is inventor and coinventor of over 12 patents. He is active in national and international scientific projects (EU, DFG, BMBF, DAAD).

Prof. Piotr Skrzypczyński

Division of Control and Robotics

Institute of Control, Robotics and Information Engineering

Faculty of Electrical Engineering

Poznan University of Technology

Poland

The aim of this tutorial lecture is to show the role of machine learning and some other AI-related techniques in embodied autonomous agents, and autonomous robots in particular.
In this tutorial we bring to the forefront the aspects of robotics that are closely related to computer science. We believe that the progress in algorithms and data processing methods together with the rapid increase in the available computing power were the driving forces behind the successes of modern robotics in the last decade. During this period robots of various classes migrated from university laboratories to commercial companies and then to our everyday life, as now everybody can buy an autonomous vacuum cleaner or lawnmower, while self-driving cars and drones for goods delivery are waiting for proper legal regulations to enter the market.
Robotics and Artificial Intelligence already went a long path of mutual inspiration and common development, starting from the symbolic AI (aka Good Old-Fashioned Artificial Intelligence) and its extensive use in early autonomous robots, such as Shakey the robot, created in SRI International by Nils Nilsson, considered one of the "fathers" of modern AI. We briefly characterize the range of the most important applications of typical AI methods in modern robotics, including motion planning algorithms [2,3], interpretation of sensory data leading to creation of a world model [4,5], and classical learning methods, such as reinforcement learning [6].
However, what made robotics a part of the new wave of AI applications was the recent "revolution" of machine learning, mostly grounded in the enormous success of the deep learning paradigm and its many variants that proved to outclass classic methods in a broad range of problems related to the processing of images and other types of signals. The quick adoption of the recent advances in Machine Learning (ML) in robotics seems to be motivated by the fact that ML gives the possibility to infer solutions from data, as opposed to the classic model-based paradigm that was for decades used in robotics. Whereas the model-based solutions are mathematically elegant and theoretically provable (with respect to stability, convergence, etc.) they often fail once confronted with real-world problems and real sensory data, as their underlying mathematical models are only a very rough approximation of the real world. Therefore, a wider adoption of ML in robotics gives a chance to make robots more robust and adaptive. On the other hand, we should try to use the new techniques without discarding the knowledge and expertise we already have - machine learning methods can benefit a lot from the prior knowledge and the known structure of the problem that has to be solved by learning. This knowledge and structure can be adopted from the model-based methods that a re already well-established in robotics. In the lecture robots are understood in a broad sense, as all embodied agents that have means to physically interact with the environment. They can be either manipulators, mobile robots, aerial vehicles, self-driving cars, and various "smart" devices and sensors.
In the second part of the lecture attention is paid to specific problems that appear in application of machine learning to embodied agents, such as the need to search a for solution in huge, multi-dimensional spaces ("curse of dimensionality"), and the ever-present problem of representation and incorporation of uncertainty in the processing of real-world data. Some examples of applications of autonomous robots are given, which were successful due to the use of AI - in particular the probabilistic representation of knowledge and machine learning. The most prominent examples are the DARPA competitions: "Grand Challenge", "Urban Challenge" and "Robotics Challenge" (DRC), and the "Amazon Picking Challenge", which proves the interest of large corporations in the development of AI-based robotics [7].
In the third part of the lecture new research directions offered by machine learning and the increased availability of training data are discussed.
An overview of the most popular application areas of ML in robotics and other autonomous systems is presented along with the typical machine learning paradigms applied in these areas. The focus is on deep learning, mostly using convolutional neural networks to process various sensory data.
We discuss three aspects of embodied agents that make machine learning in robotics quite specific with respect to other application areas, such as medical images or natural language processing.
The first aspect is dealing with the "open world", in which autonomous robots usually operate. This situation breaks the assumptions underlying some popular ML methods, and creates the need to face the problem of unknown classes identification [8] incremental learning [9], and the uncertainty of sensory data [10]. We also stress out that an embodied agent has the ability to actively acquire information [11].
The second aspect is the inference about the scene seen by the agent, where in the case of robotics, semantics and geometry intermingle [12], because the robot has to work in a three-dimensional world, although it often perceives it through two-dimensional images [13,14].
The third aspect of our analysis is related to the most important feature of robots that distinguishes them from all other learning agents (software-based). Robots are embodied agents, that is they have a physical "body", and are subject to physical constraints, such as the maximum speed of motion or maximum range of perception. Therefore, in ML for robots analysis of the spatio-temporal dependencies in data is very important [15]. Robots support advanced learning methods thanks to the possibility of interaction with the environment - a simple example is active vision with moving camera, a much more complex one is manipulation with active testing of the behavior of objects (repositioning, pushing) [16].
At the end of the lecture, in the context of specific needs and limitations characteristic to the applications of ML in robotics, new concepts of machine learning (e.g. deep reinforcement learning [17], interactive perception [18]) are presented. The lecture is summarized with a brief discussion of the most important challenges and open problems of ML applied to embodied agents.

References

[1] R. Fikes, P. E. Hart, N. J. Nilsson, "Learning and executing generalized robot plans". Artificial Intelligence, 3(4), 1972.
[2] D. Belter, P. Łabęcki, P. Skrzypczynski, "Adaptive motion planning for autonomous rough terrain traversal with a walking robot". Journal of Field Robotics, 33(3), 2016.
[3] S. Tonneau et al.. "An efficient acyclic contact planner for multiped robots". IEEE Transactions on Robotics, 34(3), 2018.
[4] S. Thrun, "Learning metric-topological maps for indoor mobile robot navigation". Artificial Intelligence, 99(1), 1998.
[5] C. Cadena et al., "Past, present, and future of simultaneous localization and mapping: Toward the robust-perception age", IEEE Transactions on Robotics , 32(6), 2016.
[6] J. Kober, J. Peters, "Reinforcement learning in robotics: A survey", In: Learning Motor Skills. STAR, Vol 97, Springer, 2014.
[7] C. Eppner et al., "Lessons from the Amazon Picking Challenge: Four aspects of building robotic systems", International Joint Conference on Artificial Intelligence, Melbourne, 2017.
[8] W. Scheirer et al.,, "Toward open set recognition". IEEE Transactions on Pattern Analysis and Machine Intelligence, 35(7), 2013.
[9] G. Csurka. "Domain adaptation for visual applications: A comprehensive survey", arXiv preprint, 2017.
[10] Y. Gal, Z. Ghahramani. "Dropout as a Bayesian approximation: Representing model uncertainty in deep learning", International Conference on Machine Learning, 2016.
[11] F. Dayoub, N. Sunderhauf, P. Corke, "Episode-based, active learning with Bayesian neural networs". CVPR Workshop on Deep Learning for Robotic Vision, 2017.
[12] D. Lin, S. Fidler, R. Urtasun, "Holistic scene understanding for 3D object detection with RGBD cameras". IEEE International Conference on Computer Vision, 2013.
[13] X. Yan, J. Yang, E. Yumer, Y. Guo, H. Lee, "Perspective transformer nets: Learning single-view 3D object reconstruction without 3D supervision". NIPS, 2016.
[14] S. Pillai, J. Leonard, "Monocular SLAM supported object recognition", Robotics: Science and Systems, 2015.
[15] N. Atanasov, B. Sankaran, J. Le Ny, G. J. Pappas, K. Daniilidis, "Nonmyopic view planning for active object classification and pose estimation". IEEE Transactions on Robotics, 30(5), 2014.
[16] D. Holz et al., "Active recognition and manipulation for mobile robot bin picking". In: Gearing Up and Accelerating Crossfertilization between Academic and Industrial Robotics Research in Europe, STAR Vol. 94, Springer, 2014.
[17] T. P. Lillicrap et al., "Continuous control with deep reinforcement learning". CoRR, abs/1509.02971, 2015
[18] J. Bohg et al., "Interactive perception: Leveraging action in perception and perception in action", IEEE Transactions on Robotics, 33(6), 2017.
[19] S. Wachter, B. Mittelstadt, L. Floridi, "Transparent, explainable, and accountable AI for robotics", Science Robotics, 2(6), 2017.
[20] D. Belter, P. Skrzypczynski, "A biologically inspired approach to feasible gait learning for a hexapod robot", International Journal of Applied Mathematics and Computer Science, 20(1), 2010
[21] L. Fei-Fei, R. Fergus, P. Perona, "One-shot learning of object categories". IEEE Trans. Pattern Analysis and Machine Intelligence, 28(4), 2006.
[22] Y. Guo et al., "Zero-shot learning with transferred samples", IEEE Transactions on Image Processing, 26(7), 2017.
[23] M. Kopicki et al., "One shot learning and generation of dexterous grasps for novel objects". International Journal of Robotics Research, 35(8), 2016.

Piotr Skrzypczyński (M:1998) received the Ph.D. and D.Sc. degrees in Robotics from Poznan University of Technology (PUT) in 1997 and 2007, respectively.
Since 2010 he is an associate professor at the Institute of Control, Robotics and Information Engineering (ICRIE) of PUT, and head of the Mobile Robotics Laboratory at ICRIE.
He is author or co-author of over 180 papers in robotics and computer science.
His current research interests include: autonomous mobile robots, simultaneous localization and mapping, multisensor fusion, machine learning, and computational intelligence in robotics.

Prof. Bart M. ter Haar Romeny, PhD

Department of Biomedical Engineering

Eindhoven University of Technology

The Netherlands

Automated, fast and large-scale computer-aided diagnosis of medical images has become reality. The greatest breakthrough is Deep Learning. It already has huge impact in self-driving cars, industrial product inspection, surveillance, robotics and translation services, and in the medical arena it is outperforming human experts already in many domains.
However, it is still largely a black box. What can we learn from recent insights in the functionality, nanometer-scale connectivity and self-organization of the human visual brain? We will discuss several recent breakthroughs in our understanding of visual perception and visual deep learning.
We apply these techniques in the RetinaCheck project, a large screening / early warning project for eye damage due to diabetes. In China now an alarming 11.6% of the population has developed diabetes, due to genetic factors and fast lifestyle changes. In this project large amounts of retinal fundus images are acquired, and the e-cloud deep learning system successfully learns to identify early biomarkers of retinal disease.
The circle is round: we can prevent blindness by learning from the visual system: vision for vision.

Bart Romeny (1952) is professor in Biomedical Image Analysis (BMIA) at Eindhoven University of Technology, the Netherlands.
1979 MSc in Applied Physics, Delft University of Technology, NL.
1983 PhD in Physics and Life Sciences, Utrecht University, NL.
1989 Associate prof. Medical Imaging, Utrecht University NL.
2001 Professor Biomedical Image Analysis, Eindhoven University of Technology, NL, emeritus per 21-12-2017.
2013 Professor Biomedical Image Analysis, Northeastern University, Shenyang, China.
His research interests focus on automated computer-aided diagnosis, and quantitative medical image analysis, using brain-inspired computing and brain network modeling. He pioneered the exploitation of multi-scale differential geometry in medical image analysis. His interactive tutorial book is used worldwide. He currently leads the RetinaCheck project, a large screening project for early warning for diabetes and diabetic retinopathy in Liaoning Province. He has developed many sophisticated retinal image analysis applications with his team, in close collaboration with clinical partners and industry.
He (co-)authored over 230 scientific papers, 14418 citations, h-index 42. He is reviewer and/or associate editor for a range of journals and conferences, and a frequent keynote speaker at conferences and summer schools.
He was president of the Dutch Society for Clinical Physics, president of the Dutch Society of Biophysics and Biomedical Engineering. He is currently president of the Dutch Society for Pattern Recognition and Image Processing, and board member of the International Association for Pattern Recognition (IAPR).
He is Fellow of the European Alliance for Medical and Biological Engineering & Science (EAMBES), and senior member of IEEE. He is recipient of the Chinese Liaoning Friendship Award in 2014. He is an enthusiastic and awarded teacher.

Prof. Heinrich Theodor Vierhaus

Computer Engineering Group

Brandenburg University of Technology Cottbus

Germany

Fault tolerant computing is a branch of technology which has developed continuously over decades since the late 1940. Application was limited to areas such as ultra-reliable computers for banks, space flight, aviation, and nuclear power stations. By that time, the extra hardware needed was a real problem that had to be accepted. Overhead often was beyond triplication, whereby extra power was acceptable. Only since about the 1990s, electronic “embedded” sub-system have found their way into new applications such automotive systems and industrial control. Typically, a robust type of electronics was employed which stayed away from smallest feature size technologies and lowest signal voltage swings. More recently, advanced features implemented by automotive electronic systems such as ultra-fast image processing towards autonomous driving strictly demand their implementation in nano-electronic technologies with a minimum feature size of 20 nm and below. Then, suddenly, there are two demands which strongly interact. ICs in nano-technologies show a rising vulnerability to disturbing influences such as particle radiation. Furthermore, the increasing stress due to scaling at constant supply voltage rather than the previous scaling at constant field strength implies a higher level on inherent stress, resulting in wear-out and shorter life times. Hence on-line fault detection and subsequent error correction becomes necessary on a wide scale. But now the extra power needed for fault tolerant computing becomes a nightmare, since extra power and extra heat promotes aging effects strongly. Research has gone two ways. First, methods of fault detection and error correction that get along with only a small extra power budget were developed. Unfortunately, they are not as powerful, robust and universally applicable as, for example, triplication plus majority vote (TMR), which consumes more than triple power. The second approach to the solution is the concept of “error resilience”. It is based on the observation that, depending on function and application of a circuit, a limited number of faults may be acceptable for some time without total system failure. If the overall systems can become aware of its own fault status, acceptance of some faults followed by a process of self-repair by re-organization during a time slot when the system is “at rest” may be a partial solution. Then only those parts of a digital system, where single or multiple bit errors will damage the function directly and critically, will need traditional “fast and hot” methods of error detection and error correction.

Heinrich Theodor Vierhaus received a diploma in electrical engineering from Ruhr-University Bochum (Germany) in 1975.
Lecturer for RF technology and electronic circuits with Dar-es-Salaam Technical College in Tanzania (East Africa) from 1975 to 1977.
Dr.-Ing. in EE from University of Siegen in 1983.
Senior researcher with GMD, the German National Research Institute for Information Technology from 1983 to 1996.
Professor for “Technische Informatik” (computer engineering) at Brandenburg University of Technology Cottbus since 1996.
He has authored or co-authored about 250 papers in the area of computer engineering, mainly related to IC test, technology, dependability and fault correction. He also contributed to three books in the area as an editor and an author. Since 2009, he has initiated and coordinated several projects on advanced education of doctoral students in the area of dependable systems, based on an international network of universities and research institutes.

Prof. Paweł Strumiłło

Medical Electronics Division

Institute of Electronics

Lodz University of Technology

Poland

Visual impairment is one of the most serious sensory disabilities. It deprives a human being of an active professional and social live. EU reports indicate that for every 1000 Europeans citizens 4 are blind or suffer from serious visual impairment and this number is predicted to increase with time due to our ageing society.
In spite of numerous, worldwide research efforts focusing on building innovative aids helping the blind no single electronic travel aid (ETA) solution has been widely accepted by the blind community. The aim of the tutorial is to apprise the current state of the art in the field of electronic interfaces aiding the blind in independent travel, navigation and access to information. Functional solutions and outcomes of recent research projects devoted to assistive technologies for the visually impaired will be presented.

Paweł Strumiłło received the MSc, PhD and DSc degrees and currently holds the position of full-time university professor at Lodz University of Technology (TUL), Poland. In 1991-1993 he was with the University of Strathclyde (under the EU Copernicus programme) where he defended his PhD thesis. His current research interests include medical electronics, processing of biosignals, soft computing methods and human-system interaction systems. He has published more than 100 frequently cited technical articles, authored one and co-authored two books. He was a principal and co-principal investigator in a number of Polish and European research projects aimed at developing ICT solutions and electronic aids for persons with physical and sensory disabilities. He received a number of prizes and awards for development of assistive technologies for the visually impaired people (in cooperation with Orange Labs). From 2015 he has been the head of the Institute of Electronics at TUL. He is the Senior Member of the IEEE and a member of Biocybernetics and Biomedical Engineering Committee of the Polish Academy of Sciences.

Prof. Adam Dąbrowski

Division of Signal Processing and Electronic Systems

Institute of Automation and Robotics

Faculty of Computing

Center of Mechatronics, Biomechanics, and Nanoengineering

Poznan University of Technology

Poland

Imaging technologies and techniques of the human eye are used for both biometric and medical-diagnostic applications. Among various types of the eye images the following can be distinguished: iris images, fundus images, and various optical coherence tomography (OCT) scans. Contemporary processing approaches to all of these image types are reviewed and analyzed together with a discussion of their applications. Advanced image processing methods and algorithms, including the artificial intelligence approach, developed at the Division of Signal Processing and Electronic Systems of the Poznań University of Technology for the considered applications, are presented. The proposed solutions are characterized by a good effectiveness and accuracy in the support of appropriate biometric and clinical decisions.

Adam Dąbrowski received a Ph.D. in Electrical Engineering (Electronics) from the Poznan University of Technology, Poznan, Poland in 1982. In 1989 he received the Habilitation degree in Telecommunications from the same university. Since 1997 he is a full professor in digital signal processing at the Faculty of Computing, Poznan University of Technology, Poland, and Chief of the Division of Signal Processing and Electronics Systems. He was also professor at the Adam Mickiewicz University, Poznan, Poland, Technische Universität Berlin, Germany, Universität Kaiserslautern, Germany, and visiting professor at the Eidgenossische Technische Hochschule Zürich, Switzerland, Katholieke Universiteit Leuven, Belgium and Ruhr-Universität Bochum, Germany. He was a Humboldt Foundation fellow at the Ruhr-Universität Bochum, Germany (1984-1986).
His scientific interests concentrate on: digital signal processing (digital filters, signal separation, multidimensional systems, wavelet transformation), processing of images, video and audio, multimedia and intelligent vision systems, biometrics, and on processor architectures. He is author or co-author of 5 books and over 500 scientific and technical publications. Among them he is one of the co-authors of "The Computer Engineering Handbook" (first edition in 2002, second edition in 2008) bestseller and most frequently cited book of the CRC Press, Boca Raton, USA.