Digital Water

Organisers:

• Nikola Milivojević, Jaroslav Černi Institute, Belgrade, Serbia
• Boban Stojanović, Department of Mathematics and Informatics, Faculty of Science, University of Kragujevac, Kragujevac, Serbia
• Milan Stojković, The Institute for Artificial Intelligence Research and Development of Serbia, Novi Sad, Serbia; Faculty of Civil Engineering, University of Novi Sad

Water is one of our most critical resources, and its management today requires innovative approaches to address mounting challenges such as climate change, urban growth, and resource sustainability. In recent years, the concept of "Digital Water" has gained increasing significance, aiming to boost the transformative impact of advanced digital technologies on water management, thereby ensuring greater resilience and efficiency in increasingly complex conditions.

This session aims to highlight how technologies such as the Internet of Things (IoT), Artificial Intelligence (AI), Machine Learning (ML), and Digital Twins are driving the digital revolution in the water sector. These technologies enable the collection of vast amounts of data, allowing for a deeper understanding of complex water systems, faster and more informed decision-making, and proactive management. Instead of relying on reactive approaches, water utilities are now moving towards predictive and condition-based strategies, which lead to improved efficiency, reduced water losses, and more reliable service delivery for customers.

Participants in the session will gain new insights into how digital transformation can be applied to the water sector from both practical and strategic perspectives. Through insights into successful implementations, including examples from domestic and international projects, participants will understand how these technologies can support their work in research, engineering, or water utility management.

The session will cover the role of IoT sensor networks in water systems, with a focus on their integration within distribution networks for continuous monitoring of key parameters. The session will also address the application of Artificial Intelligence and Machine Learning in tackling issues related to water resource management. In addition to these topics, there will be a focus on the implications of cybersecurity in the context of digital water infrastructure. As the water sector increasingly integrates IoT and other digital solutions, it becomes more vulnerable to cyber threats, which means ensuring the resilience of digital water systems involves addressing these security risks directly.

Relevant topics include (but not limited to):

• Digital Technologies in Water Monitoring and Management (IoT and ML Technologies for Water System Monitoring, AI in Real-Time Monitoring and Prediction of Water Quality, ML Models for Flood Prediction, Digital Twins in Water Management)
• Advanced Analytics and Optimization in Hydrological Systems (Hybrid Physically-Based and ML-Based Models, AI and ML Supported Reservoir Operation Optimization, Integrative Data Analytics)
• Nature-Based Solutions and Sustainability (AI-Informed Nature-Based Solutions, Digital Tools for Improving Aquatic Environments)
• Decision Support, Policy, and Governance (Decision-Support Systems for Water Resource Planning, Policy and Governance in Digital Water Management, Case Studies on Successful Digital Transformations in the Water Sector)
• Security and Innovative Solutions (Cybersecurity in Digital Water Systems, Innovative Digital Solutions for Water Systems)
• Robust and Trustworthy AI under climate extremes (Uncertainty quantification, reliability assessment and robustness under extreme conditions)

ICT for health, well-being and sport

Organisers:

• prof. Nenad Filipović. Faculty of Engineering, University of Kragujevac, Kragujevac, Serbia BIOIRC Bioengineering Research and Development Center, Kragujevac, Serbia
• prof. dr Anton Kos. Faculty of Electrical Engineering, University of Ljubljana, Slovenia, Ljubljana, Slovenia

One of the greatest social challenges today is addressing many different concerns and issues related to healthcare and well-being. These challenges include, but are not restricted to, improving the understanding of the causes and mechanisms connected to health; ability to monitor, prevent, detect, and treat disease; personalize health and care delivery. For addressing such challenges, a new approach is needed, which will transform the health and care delivery. This approach is proactive rather than reactive. Healthcare should evolve mechanisms for early risk detection, shifting from the acute treatment to long-term health monitoring. Instead of focusing on single disease treatment, healthcare systems should become capable of handling multiple chronic conditions and comorbidities at the same time. Care delivery should become pervasive, not exclusively institutional. Finally, the generic treatment approach should be abandoned, and care should become more personalized.

Although health is an important, if not necessary, component of well-being, physical activities such as recreation and sports are a good way to maintain and improve it. With the rapid advancement and affordability of miniature, wearable sensor devices, new approaches to quantifying and qualifying our physical activities are possible. For example, innovative applications based on sensor data will contribute to faster, cheaper and safer physical rehabilitation programs, personalized recreational plans, and in-depth analysis of athletes' performance in different sports.

Advanced ICT technologies will play a crucial role in this transformation. Health-related data will grow and will have accurate and useful digital representations. Predictive computing, used by the care providers and customers, will use health-related data to facilitate early risk detection and self-management of health and disease. Health systems will become interoperable. Robotic systems, driven by artificial intelligence will enable assisted living, in smart, age-friendly homes. Biofeedback applications in healthcare, sport, recreation, and rehabilitation will enable an easier and faster way to achieving the desired results. They will assist individuals, healthcare personnel, and trainers in many ways; from influencing the physiological processes in the human body, to motor learning in sports and rehabilitation.

With this session, we aim to uncover the current research in the above mentioned topics and topics closely related to them. All relevant contributions are strongly welcomed.

AI & IoT for Smart Industry

Organisers:

• Milovan Medojević, The Institute for Artificial Intelligence R&D of Serbia, EnergyPulse DOO Novi Sad
• Aleksandar Rikalović, Faculty of Technical Sciences, University of Novi Sad, Novi Sad, Serbia
• Milan Zdravković, Faculty of Mechanical Engineering, University of Niš, Niš, Serbia

The mutually beneficial relationship between the internet of things (IoT) and Artificial Intelligence (AI) enables disruptive innovations in industry worldwide where the fusion of AI and IoT enables the systems to be predictive, prescriptive, and autonomous. This convergence of AI and IoT is evolving while its continuum impacts variety of industries ranging from manufacturing, retail, healthcare, telecommunication, and transportation, etc. In this context, IoT solutions and real-time data processing can empower a massive amount of data captured from interconnected devices, while Big Data Analytics and AI-based solutions can help tackle many concerns and achieve intelligent prediction, evaluation, optimization, and decisionmaking. However, various challenges like interoperability, decentralization, distributed control, real-time process control, service orientation, and maintenance optimization are being addressed also.

Additionaly, numerous challenges exist in implementing such systems that include algorithmic and design innovations to meet Quality of Service requirements (latency, bandwidth, delay, etc); mechanisms to preserve IoT data privacy and provide secure services for interconnected users; achieving high performance systems that can process both high volume and fast speed IoT data leveraging Edge AI.

For the aforementioned reasons, this research topic aims to solicit the submissions of original and unpublished research articles that present in-depth fundamental research contributions either from theoretical or methodological/application perspective containing novel architectures, algorithms, systems, techniques or applications offering new insights and findings in the field of AI-powered IoT for smart industry of tomorrow.

We seek high-quality submissions related to (but not limited to) one or more of the following topics:

• Big Data Analytics for Industry 4.0
• Artificial Intelligence and Evolutionary Techniques for Industry 4.0
• AIoT for Industry 5.0
• Intelligent Monitoring & Control Systems
• Data-driven Modelling and Optimization
• Machine & Deep Learning for Intelligent Systems
• Data Analytics for Scheduling Industry Processes
• Data Analytics for Monitoring and Control of Industry Operations
• Computational Intelligence Technologies for Industry
• Cyber-physical systems
• Real-time condition monitoring in manufacturing
• Machine learning and deep for fault diagnosis
• Planning and scheduling in Industry 4.0
• Advances in smart manufacturing
• IoT systems for smart manufacturing
• Digital twin for smart manufacturing
• Industrial big data and data analytics
• Manufacturing cyber security
• IoT and AI techniques
• Potential of AI-Based Automation for IoT-Enabled Smart industries
• Machine learning and AI for IoT data processing and analysis
• Edge AI for industry-centric IoT systems and human-machine interaction
• New Applications for smart IoT Services
• Exploring the Role of AIoT in Smart Energy Management
• Small to Large-scale AIoT pilots
• IoT testbeds and testing tools
• Predictive Maintenance and AIoT
• Smart Manufacturing, Energy, Environment & Water Management and IoT
• Smart Security and AIoT
• Smart Metering and AIoT
• Smart Analytics and AIoT
• AI for Financial systems
• Smart Logistics and AIoT

Generative AI and Large Language Models (LLMs)

Organisers:

• Prof. dr Aleksandar Stanimirović (Faculty of Electronic Engineering, University of Niš, Serbia)
• Prof. dr Miloš Bogdanović (Faculty of Electronic Engineering, University of Niš, Serbia)

Generative AI, including Large Language Models (LLMs), Generative Adversarial Networks (GANs), Diffusion Models, and Neural Radiance Fields (NeRFs), has revolutionized content creation across domains such as text, images, 3D models, and even video. With the rise of LLMs like GPT, BERT, Llama and Mistral, alongside powerful generative models like DALL-E, Midjourney, and Stable Diffusion, the impact on industries as diverse as entertainment, healthcare, education, and design is profound. But the rapid advancement of these models also brings significant engineering challenges, from optimizing architectures and training methods to ensuring scalability and robustness.

This special track aims to explore recent technical innovations, practical applications, and unresolved challenges in the field of Generative AI. We invite researchers and engineers to submit their work related to the design, optimization, and deployment of these models, with a particular focus on enhancing efficiency and reliability of these systems.

Topics of Interest

• Innovative architectures for generative models, such as advancements in Transformer-based models or diffusion models for image and video generation.
• Optimization algorithms that enhance the performance, efficiency, and scalability of Generative AI models.
• Model compression and transfer learning techniques for reducing the computational load of Generative AI.
• Efficient training techniques for large-scale models, including distributed training, parallelism, and memory optimization for better resource utilization.
• Zero-shot and few-shot learning for generative tasks, enabling models to generate high-quality content from minimal data.
• Multimodal AI techniques that integrate text, images, audio, and video for comprehensive content generation.

• Innovative applications of Generative AI in various domains, such as medicine, education, business, finance, etc.
• Advances in models that convert natural language into high-fidelity images and videos, improving content creation for entertainment, design, and marketing.
• Educational technologies leveraging Generative AI, such as chat-bots, content generation, feedback systems, etc.
• Techniques for generating synthetic datasets for training other machine learning models, especially in healthcare, autonomous driving, and robotics.
• Generating synthetic medical images, simulations, and data for training diagnostic systems and aiding in drug discovery.

• Ensuring that generative models can withstand adversarial inputs and produce reliable outputs in high-stakes applications.
• Developing techniques that give users control over the behavior and outcomes of generative models, particularly in ensuring adherence to constraints like style or factuality.
• Addressing issues where LLMs generate incorrect or fabricated information, with a focus on improving the reliability of outputs in sensitive fields.
• Addressing the challenge of limited high-quality data to enable LLMs to learn and perform effectively in languages with fewer available digital or labeled resources.
• Overcoming the high computational costs of training and deploying large generative models, with a focus on improving energy efficiency and scaling across distributed systems.
• Engineering solutions to enable real-time output from generative models, especially in applications such as gaming, VR, and live content creation.
• Developing standardized metrics and tools to assess the performance, efficiency, and scalability of generative models across diverse tasks and domains.