Fuzhou University, China

An Investigation on the Influence of Blood Volume in the Cardiac Cycle on Channel Gain of Intracardiac Communication Channels

Cardiovascular diseases are the leading cause of death worldwide. According to statistics from WHO, approximately 18 million people died from cardiovascular diseases in 2016, accounting for 31% of all disease-related deaths. Currently, pacemaker treatment is the most effective method for treating cardiac arrhythmias. However, in patients implanted with conventional pacemakers, there is always a risk of electrode displacement or breaking, which can lead to complications such as hematomas and infections. Leadless cardiac pacemakers (LCP) have been proposed to overcome these problems. Because LCPs have no electrodes, surgery is less invasive and time-consuming and the potential complications that arise from the use of electrodes are eliminated. Multi-chamber pacing requires communication between LCPs within the heart so that the different LCPs in the different chambers can operate sequentially.

In the galvanically coupled conductive intracardiac communication (GCCIC) of the leadless pacemaker, the electrical signal transmitted directly through the myocardium and blood is inevitably affected by the cardiac cycle. Previous studies have focused more on the effects of the myocardium on signal transmission. However, our preliminary in vitro experiments suggest that variations in blood volume also have a significant impact on signal transmission. In this presentation, we analyzed the blood volume variations during the cardiac cycle and designed an in vitro experimental platform that included a simulated heartbeat system and an automatic channel gain acquisition system that controls two peristaltic pumps to realize the periodic blood volume variations and continuous channel gain acquisition. Through the in vitro porcine heart experiment, the effects of frequency and blood volume variations during the cardiac cycle on the gain of the two channels were analyzed.

Biography

Dr. Yue-ming Gao received the Ph. D. degree in electrical engineering from Fuzhou University, Fuzhou, Fujian Province, China, in 2010. He is now a professor and head of the Department of Electronic Information Engineering, College of Physical and Information Engineering, Fuzhou University, and the Associate Director of the International Science & Technology Cooperative Center of Health & Medical Instrumentation, Chinese Ministry of Science and Technology (MOST).

His research interests include biomedical signal detection and processing, biomedical instrumentation, traditional Chinese medicine engineering, body area networks, wearable and implantable healthcare devices, photoelectrical readers for point of care testing (POCT).

He is now IEEE Senior Member (EMBS), Senior Member of Chinese Institute of Electronics (Biomedical Electronics Branch), Senior Member of Chinese Society of Biomedical Engineering, Vice Chair of Biomedical Engineering Society of Fujian Province, China, as well as Associate Editor of IEEE Journal of Electromagnetics, RF and Microwaves in Medicine and Biology.

TU Berlin, Germany

Semantic Streams for Cooperative Perception

In this talk, I will present how to use semantic streams to model and build perception systems from multi-modal sensory stream data.  Semantic streams mimic the semantic memory and episodic memory of the human cognitive system. The semantic memory refers to our brain’s repository of general world knowledge. Whereas the episodic memory refers to our “episodic memory system”, which encodes, stores, and allows access to “episodic memories”, e.g. recollection of personally experienced events situated within a unique spatial and temporal context. In this context, semantic and episodic memories are represented as semantic and stream graphs to fuse and capture various kinds of sensory observations, e.g, images, videos and point clouds, into interlinked sub-symbolic and symbolic data streams at different levels of semantic abstraction. 

During the talk, I will walk through the process of building some data fusion components for autonomous driving and robotics via a declarative programming model based on semantic streams. This programming model enables developers to write semantic stream reasoning programs which are composed from if-then rules associated with stream data fusion operations for both reasoning and learning tasks. As our system is under active development, I will share some engineering experiences and lessons learnt to facilitate collaborations and to seek for advice as well as new ideas.

Biography

Danh Le Phuoc is a pioneer in the area of semantic stream processing and reasoning. He built some first systems in RDF stream middleware and processing which won various awards in the Semantic Web community, e.g, honorable test-time award. He is also known for his work as a co-editor of the W3C/OGC standard, Semantic Sensor Network Ontology. These systems and standards have been used in various EU, Irish and German projects, e.g COSMO, SMARTER, SMARTEDGE and AIoTwin. Among them, he is the technical coordinator of the EU Horizon project, SMARTEDGE (Semantic Low-Code Programming Tools for  Edge Intelligence), with 15 partners including Bosch, Siemens, Dell and NVIDIA to build emerging cooperative systems such as collaborative robots in Dell and Siemens factories and V2I and I2V systems with Helsinki city and Bosch. 

He is currently a research group leader at Berlin Institute for Foundation for Data and Learning (bifold.berlin) at TU Berlin where he is leading the research lab, ‘Pervasive Intelligence and Computing’, PICOM.AI. The research interests of PICOM Lab are around the intersection of knowledge graph, neural-symbolic AI and pervasive computing with the applications for autonomous vehicles, robotics and industrial IoT. Besides, he is also a senior scientist at Fraunhofer FOKUS.

Ororatech, Germany

Opening up space for Earth observation to non-initiates

Distrust on "official" sources has increased the public interest on monitoring human and economical activities, specially regarding its environmental effects. At the same time, there is a notorious increase in the frequency and intensity of natural disasters affecting larger areas of the globe. Technology is now mature enough to offer remote sensing capabilities at a fraction of the cost and complexity than 1 or 2 decades before. Depending on the area of coverage required and the time intervals, space based earth observation might be the only viable option. But here too, technological advance has made the field open for technically sound, but non-governmental, groups. There is a wide set of companies that are performing here successfully, applying state-of-the-art technologies. The next step should be now to lower the entry-barrier for less space-experienced groups, both on space-related complexity and on economical effort. National space agencies and ESA have recognized this trend and started efforts to provide options, tools and financing to interested parties, coronated finally in the form of In Orbit Demonstration or Verification (IOD/IOV) missions.

Biography

Pablo received his Dipl.-Ing. in Mechanical Engineering at ITBA in Buenos Aires, and his Dr.-Ing. degree also in Mechanical Engineering from the Karlsruhe Institute of Technology (KIT), working on the simulation of particular mechanisms. He has later gained more than 5 years of experience in small satellite development projects at Satellogic in Argentina as one of their first employees, leading the mechanical, thermal and propulsion development of 4 successful earth observation satellite missions there, comprising 5 spacecraft in total. These satellites had an increasingly higher resolution optical EO-focus, reaching around 1.5m GSD with an MTF higher than 15%. In January 2022 he joined Ororatech in Munich, Germany, to help develop novel themal-IR sensing satellites tasked with monitoring Wildfires and other events showing strong thermal contrast.

Pablo remains involved in academic activities at ITBA since graduating there, participating in more than 6 different courses, having led more than 20 theses and published around 10 papers and conference presentations. Since 2019 he is participating on deep reinforcement learning projects at the university as a side activity.

Eindhoven University of Technology, Netherlands

Resilient Power Electronics Systems for the Energy Transition

As the world is moving towards a more sustainable use of energy in general, and electricity in particular, energy processing systems (EPSs) based on power electronics (PE) technology are in the heart of a far-reaching transformation that spans across many different applications and power levels, supporting the electrification of a wide range of human activities. A key enabler for many of the changes under way, PE technology is also however heavily affected by them, as requirements for high efficiency, increased productivity and flexible operation become ever more demanding. Underpinning these requirements is the need for the new EPSs to operate reliably, and demonstrate resilience by maintain their availability for the users even in the presence of some faults or component failures. But how do these systems fail and what reliability challenges are they facing in industrial applications today? And what opportunities for increased resilience arise from the advances in computational intelligence that is either embedded in today’s EPSs or easily accessible to them by exploiting the available modern connectivity options? In this talk we will provide an overview of these questions and attempt to showcase some of the possibilities that exist.

Biography

Georgios Papafotiou obtained his PhD in Electrical Engineering, focusing on Automatic Control and Power Electronics Systems, from the Aristotle University of Thessaloniki (Greece) in 2002. From 2002 to 2006 he was with the Automatic Control Laboratory at ETH Zurich (Switzerland), where he worked on the application of Model Predictive Control (MPC) for Power Electronics Systems. From 2006 to 2021 he was with ABB. There, he held various positions, starting from Scientist and Principal Scientist to Group Leader of the Power Electronics team in the Swiss Lab of ABB Corporate Research, to Global R&D Manager for SW and Control HW with Medium Voltage (MV) Drives and finally Quality and Customer Experience Manager for the System Drives Division. Since 2021, Papafotiou is a professor in the Department of Electrical Engineering, and the Head of the Power Electronics Lab at TU/e.

University of Oxford, United Kingdom

Data-driven algorithms for multi-agent optimization and games

The advances in information and communication technology have led to the development of smart infrastructures, whose size and complexity constitute a major challenge in control engineering. Smart infrastructures notably comprise a variety of service providers and interconnected users that interact and share resources. Such a network of "ecosystems", each with high levels of autonomy, pose questions on how to better coordinate their actions, while ensuring a provably safe, efficient and uninterrupted operation? In this talk, we will show how the frameworks of multi-agent optimization, game theory, and statistical learning can be leveraged in synergy for the modelling and synthesis of solutions.

Biography

Dr. Filiberto Fele is Research Associate with the Energy and Power Group and the Control Group, Dept. of Engineering Science, University of Oxford. He has been involved in two ambitious projects set in the context of energy systems transition to zero carbon, namely V2GO (vehicle-to-grid Oxford) and LEO (Local Energy Oxfordshire). His experience with these projects highlighted the complexity and uncertainty of the supply-demand interplay at the grid edge. As these drive the adoption of technical solutions based on data-driven techniques, it is critical to ensure that datasets used for decision making form exhaustive descriptions of the observed system. Hence, his current work is dedicated to accompanying data-based methods with rigorous certificates regarding their performance in presence of unobserved behaviours, to ensure the robustness, stability and safety of the grid.

Igor Sikorsky Kyiv Polytechnic institute, Ukraine

DC-DC converters of hybrid energy storage systems for pulsed load application

The vast majority of portable devices and systems have complicated and nonlinear power consumption characteristics. The power supplies of these devices and systems, as well as the energy storage of the Microgrid and wind or solar power generation, have to provide average and peak load power, required weight and size, high energy efficiency. Various battery types are typically utilized in these power supply due to the high energy density of the cell. However, the high peak load current that is significantly greater than the nominal battery discharge current could lead to battery life cycle reducing and deteriorating cell characteristics. The hybridization of high power density storage and high energy density storage could be an effective solution to the problem of power supply portable devices and energy storage in Microgrid, solar and wind energy systems.

Thus, the implementation of hybrid energy storage systems based on batteries and supercapacitors for self-contained power supplies and portable systems is becoming a new promising research field in recent years. Another perspective area of battery supercapacitor energy storage application could be power supplies of a micro resistance welding equipment, that also have pulsed power consumption. Energy storage that is simultaneously characterized by high power and energy density is required for micro resistance welding power supplies based on capacitive discharge topology. Auxiliary DC-DC power converters are used for energy distribution between battery and supercapacitor. This work focuses on the DC-DC converter of hybrid energy storage that is used in a system with pulsed current consumption on the example of welding current for micro resistance welding technology.

Biography

Yuliia Kozhushko received the B.Sc. and M.Sc. degrees in electrical engineering from the National Technical University of Ukraine Igor Sikorsky Kyiv Polytechnic Institute, Ukraine, in 2015 and 2017, respectively. Her experience includes five years of practical engineering work as an Embedded Developer. In the 2021 she received the Ph.D. degree, with a dissertation devoted to the research of hybrid energy storage systems for the pulsed load application, from the Igor Sikorsky Kyiv Polytechnic institute. He is the author or coauthor of six scientific articles. Her research interests include design, simulation of power electronic converters, and control systems, hybrid energy storage systems, resistance micro welding.

University of Zagreb Faculty of Electrical Engineering and Computing, Croatia

Towards Intelligent Soft Robotics

In the coming decades we will need to dramatically increase work productivity not only to cope with a shrinking work-force and growing number of people in old and very old age in developed countries, but also to mobilize resources to help the ecologically sustainable development of the global economy and provide food and infrastructure to billions of more people. A steep progress in Robotics and AI seems a dramatic necessity in this context. Luckily, the current wave of innovation in Robotics, integrating Machine Learning, Probabilistic Robotics, and some AI is already having significant impact on our economy and our society. However, it is a widespread opinion that Robotics still need much more robustness, safety, lower manufacturing costs, and reduced control complexity and effort, while it aims to more and more complex and adaptive behaviors in open-ended environments. This talk will describe a sensible strategy to harvest the low-hanging fruits and making significant progress to cope with the still open foundational issues. To what extent, under which conditions – and critically in which timeframe - will it be possible to achieve the necessary advancements to be able to exploit robots to dramatically increase productivity and for elder care? Which mathematical challenges will arise? Are those bottlenecks to AI and robotics development?

Biography

Prof. Fabio Bonsignorio is the ERA Chair in AI for Robotics and head of the AIFORS Lab, in LAMOR at FER at the University of Zagreb. AIFORS lab aims to take intelligent soft robotics to the next stage. He is CEO and Founder of Heron Robots (advanced robotic solutions). He has been Visiting Professor at the Biorobotics Institute of the Scuola Superiore Sant’Anna in Pisa in the period 2014-2019. He has been professor in the Department of System Engineering and Automation of the University Carlos III of Madrid until 2014. In 2009 he was awarded the Banco de Santander Chair of Excellence in Robotics at the same university. He has been working in the R&D departments of several major Italian and American companies, mainly in the applications of intelligent systems and technology transfer with coordination/management responsibilities for more than 20 years.

He is a Founding Director of euRobotics aisbl, the private part of SPARC, and now part of ADRA. He is a past elected member of the Research Board of Directors of SPARC. He coordinated and has been the main teacher of the ShanghAI Lectures since the 2013 edition. The ShanghAI Lectures are an advanced network MOOC teaching initiated several years ago by Rolf Pfeifer. He has pioneered and introduced the topic of Reproducible Research and Benchmarking in Robotics and AI, where is one of the leading experts, if not the leading one. He is a pioneer in the applications of Blockchain technologies in Robotics.

He is a senior member and a Distinguished Lecturer of IEEE/RAS . He coordinated the EURON Special Interest Group on Good Experimental Methodology and Bench- marking in Robotics, is Co-Chair of the IEEE RAS TC-Pebras and has been a board member of EURON III. He is a member of the GeorgeGiralt PhD Award jury. He is the coordinator of the euRobotics Topic Group on Experiment Replication, Benchmarking, Challenges and Competitions and is co-chair of the IEEE TC- Pebras. He has participated to the design and launch the euCognition society. He is in the Management Committee of the COST Action CA17137 - A network for Gravitational Waves, Geophysics and Machine Learning, whose has been one of main proposers. He co-organized the 1st Conference on Machine Learning for Gravitational Waves, Geophysics, Robotics, Control Systems.

Jožef Stefan International Postgraduate School, Slovenia.

The dynamics of leaders and followers during physical collaboration in human-robot dyads

Recent technological developments allow robots to safely share a common workspace with humans. Europe is currently leading the market for safety-certified robots, providing tools that can respond to unintended contact. However, robots still have a shortcoming when it comes to physical collaboration between humans and robots due to their limited ability to observe human dynamics. This leads to inefficient collaboration and unergonomic interaction. In my talk, I will present the work of my Laboratory for Advancing Collaborative Robot Behavior in Physical Human-Robot Interaction Scenarios (CoBoTaT), which aims to leverage these technologies by providing tools that can efficiently observe human dynamics in real-time by learning predictive models from datasets and incorporating these models into online robot control to make collaboration more efficient.  

Biography

Tadej Petrič is currently a senior research associate in the Department of Automatics, Biocybernetics and Robotics at the Jožef Stefan Institute and an assistant professor at the Jožef Stefan International Postgraduate School, Slovenia. He is also the head of the Laboratory for the Advancement of Collaborative Robot Behaviour in Physical Human-Robot Interaction (CoBoTaT). In 2015, he was a postdoctoral fellow at Biorob (Prof. Auke Ijspeert lab) at École polytechnique fédérale de Lausanne (EPFL). In 2013, he received the D.Sc. degree in Robotics from the Faculty of Electrical Engineering, University of Ljubljana. He conducted part of his doctoral research in the Department of Robotic Systems for Dynamic Control of Legged Humanoid Robots at the German Aerospace Center (DLR) in Germany. In 2013, he was a visiting researcher at ATR Computational Neuroscience Laboratories in Japan. His current research focuses on the development of biologically plausible robot controllers that achieve robustness and adaptation to changing environments comparable to humans.

eCampus University, Italy

Implementation of a Deep Convolutional Neural Networks model for automated diagnosis of COVID-19 pulmonary diseases using radiographic images

Today's diagnostic imaging technology used to detect diseases of the human body makes available many high-quality, pre-classified and open datasets. The images acquired are to be interpreted for diagnosis, prognosis and disease treatment planning. Understanding such medical images is generally performed by experienced medical professionals, however, their availability and the effort required often limits their effectiveness. The COVID-19 outbreak is having a devastating effect on global well-being and public health. More than 604 million confirmed cases have been reported worldwide until now. Due to the continuously growing number of confirmed cases, and challenges to the variations of the COVID-19, timely and accurate classification of healthy and infected patients is fundamental in order to control and treat COVID-19. We developed from scratch a deep learning-based system based on Deep Convolutional Neural Networks (DCNN) for the persuasive classification and reliable detection of COVID-19 using chest X-Ray images for training and validation purposes. The resulting trained model is able to classify, with an accuracy of 94. 01%, the images between those with suspected pulmonary pathology related to COVID-19 infection, other pneumonia (bacterial or viral) or normal (no evident pathology) and those with tertiary classification distinguished as 'covid', 'normal', 'pneumonia'; this accuracy was also confirmed at 93.23% by evaluating the trained model also with a different dataset of 9208 validated and publicly available posteroanterior chest X-ray images from the Mendeley Data site.

Biography

Prof. Cristian Randieri is a great proactive person and an effective communicator and a visionary. His knowledge is not restricted to the area he covers, but his impressive passion for all the technical topics led him to acquire great skills even in areas far from his original studies. He is a valuable writer with a great scientific background formed with more than 15 years of active research on experimental Nuclear Physics performed in the most famous international research laboratories such as CERN, ESRF, INFN. He writes about HI-Tech solutions topics and advanced research studies applied to the industry. With more than 200 scientific and technical publications, he is also, a well-known technical writer in Italy because of his interviews published in the most famous Italian industrial magazines. He is the Founder of Intellisystem Technologies an Italian Research & Developments company committed to develop and sell innovative and advanced solutions. Also, he serves as external reviewer for the NASA Postdoctoral Program (NPP) at USRA and for the NASA MSI Fellowship. Actually, he is a reviewer of the Heliyon International Journal, and an Adjunct Professor at eCampus University, where he teaches Computer Vision, Data Base and Human Machine Interfaces.