|
 |
Hybrid Event
|
Papers |
P. Lončar, P. Lončar (University of Split, Split, Croatia) Evolutionary Computation Algorithms for Task Scheduling in Distributed Computing 
Numerous metaheuristic algorithms have been used to optimize resource allocation challenges in heterogeneous distributed computing systems, such as heterogeneous cloud environments. One such challenge is assigning tasks to available resources based on specific criteria to reduce execution time, improve throughput, and enhance system performance. The research revealed that the majority of studies were based on Computational Intelligence algorithms inspired by natural systems. In this paper, four algorithms from the Evolutionary Algorithms group are evaluated under different loads to provide a comprehensive insight into their performance by applying simulation-based optimization as a research method. The experimental results of our implementation of the Genetic Algorithm, Differential Evolution, Evolution Strategies and Evolutionary Programming are compared considering several performance metrics. Overall results demonstrate that the Evolution Strategies algorithm exhibits the best performance. Potential future research directions and associated challenges are discussed as well.
|
R. Piltaver, A. Janach (Fakultet informatike i digitalnih tehnologija - Sveučilište u Rijeci, Rijeka, Croatia) A Decision-Making Framework for Evaluating Orchestration Systems in Microservices Applications 
This paper introduces a decision-making framework for the systematic evaluation and ranking of orchestration systems for microservices applications. Key evaluation parameters such as performance, cost and management complexity are identified based on a comprehensive analysis of orchestration systems and used to develop a qualitative multi-attribute decision model in DEXi. The model is applied to two orchestration systems, Azure Kubernetes Service (AKS) and K3S, and tested against the needs of two hypothetical companies with distinct operational profiles. Results demonstrate how the model facilitates nuanced decision-making by balancing organizational priorities with the technical and operational capabilities of orchestration systems.
|
F. Asani, J. Ajdari, X. Zenuni (Faculty of Contemporary Sciences and Technologies, Tetovo, Macedonia) Agentic AI for Dynamic Cloud Resource Management: A Conceptual Framework for Dynamic Cloud Resource Management 
Effective cloud resource management is crucial for modern computing, yet traditional rule-based methods fail to adapt to rapidly changing workloads. While AI techniques like reinforcement learning, predictive analytics, and multi-agent systems offer promising solutions, they often operate in isolation, lacking intentionality, collaboration, and autonomy required for comprehensive resource management in dynamic environments.
This research explores Agentic AI, a novel approach that integrates autonomous, goal-driven, and collaborative agents to optimize resource allocation in cloud-native systems. Through a systematic review of existing literature, this study identifies critical gaps in current techniques and proposes a conceptual framework leveraging Agentic AI principles. The framework enables agents to predict workloads, set system-wide goals, and dynamically adjust resources in real-time.
By addressing these gaps, this research contributes a unified, adaptable solution to enhance scalability, workload distribution, and cost-efficiency in distributed cloud environments, advancing the state of intelligent cloud resource management.
|
N. Nelufule, D. Shadung (Council for Scientific and Industrial Research - CSIR, Pretoria , South Africa) Exploring the Role of Dew Computing in Enhancing Cybersecurity and Digital Forensics: A Systematic Literature Review 
The Fourth Industrial Revolution has precipitated the emergence of sophisticated cyberthreats, which evolve with time. The evolution of these threats demands innovative solutions applied in cybersecurity and digital forensics to improve data security and incidence response. This paper presents a systematic literature survey conducted on dew computing, which integrates local computing capabilities with cloud resources to create a hybrid framework that enhances data accessibility, security, and forensic analysis. By allowing devices to operate independently of constant internet connectivity, dew computing addresses significant challenges faced by traditional cloud computing, such as latency and data vulnerability during transmission. The main objective of this paper was to study how dew computing can enhance real-time, enhance data protection and data integrity, and enhance digital evidence acquisition. Furthermore, it discusses the challenges and limitations associated with implementing dew computing, including technical barriers and privacy concerns. The research survey findings suggest that dew computing offers a promising approach to mitigating cybersecurity risks and provides digital forensic investigators with the tools necessary for effective investigations.
|
J. Vehovar, M. Rot (Institut "Jožef Stefan", Jožef Stefan International Postgraduate School, Ljubljana, Slovenia), G. Kosec (Institut "Jožef Stefan", Ljubljana, Slovenia) Performance trade-offs of high order meshless approximation on distributed memory systems 
Meshless methods approximate operators in a computational node as a weighted sum of values in its neighbours. Higher order approximations of derivatives provide more accurate solutions with better convergence characteristics, but they come at the cost of including more neighbours. On the accuracy-per-compute time basis we know that increasing the approximation order is beneficial for a shared memory computer, but there is additional communication overhead when problems become too large and we have to resort to distributed memory systems. Computational nodes are divided between systems in spatially coherent subdomains with approximations at their edges requiring neighbouring value exchange. Performance optimization is then a balancing act between minimizing the required number of communicated neighbours by lowering the approximation order or increasing it to enable faster convergence. We use the radial basis function-generated finite difference method (RBF-FD) to approximate the derivatives that we use to solve the Poisson equation with an explicit iterative scheme. Inter-system communication is provided by Open MPI, while OpenMP is used for intra-system parallelisation. We perform the analysis on a homogenous CPU-based cluster where we examine the behaviour and attempt to determine the optimal parameterisation with the goal of minimizing the computational time to reach a desired accuracy.
|
S. Vojvodić, K. Vidović, D. Leljak, A. Najev (Ericsson Nikola Tesla d.d., Zagreb, Croatia), N. Grubišić (Faculty of Maritime Studies, University of Rijeka, Zagreb, Croatia) A Traffic Decision Support System Assisted by Mobile Network Data 
A traffic decision support system based on an integrated architecture combines a digital traffic model with real-time data processing capabilities. It provides estimates of traffic demand in specific areas and can forecast future traffic states. The model ingests a continuous stream of different realtime traffic data from various sensors, including vehicle counts, speed, direction, and type. Additionally, two types of mobile network data are used as support input. First, anonymized mobile data offering insights into mobility patterns across the area of interest within a selected timeframe. This data is reliable, objective, and consistent, providing more comprehensive and statistically relevant information than traditional survey-based data. Moreover, it gives valuable insights into macro and meso-scale migration patterns. Second, statistical data on mobile network load allows for real-time assessments of traffic density and population estimates in specific areas. By aggregating these different sets of real time data and implementing specific mobile network data for model improvement, the proposed system can create a digital twin of the traffic network able to emulate and predict the current real world traffic situations. Results of traffic modelling and data analysis yield operational support and can be used for both current traffic solutions and future planning.
|
A. Cvetković, M. Miletić, T. Špoljarić, G. Malčić (University of Applied Sciences, Zagreb, Croatia) Parking Lot Availability IoT System Based on ESP8266 Platform 
An electronic system for parking lot availability based on ESP8266 microcontroller is described in this article. Components of the system include: a solar panel with rechargeable 18650 battery, an ESP8266 microcontroller with wireless communication and a magnetometer sensor. MQTT protocol is being used as communication for efficient data transfer between the parking sensors and the central server. Also, a web interface is developed for real-time parking lot tracking, which provides easy information accessibility on multiple platforms.
|
Break |
Papers |
A. Shekerov (Ss. Cyril and Methodius University in Skopje, Skopje, Macedonia), A. Angjelevska (Innovation Dooel, Skopje, Macedonia), M. Gusev (Ss. Cyril and Methodius University in Skopje, Skopje, Macedonia) Efficient Time-Series Heart Rate Variability Metrics in C++ 
The focus of this paper is the efficient computation of Time-Series Heart Rate Variability (HRV), an important set of metrics in modern medicine and cardiology. HRV metric calculation poses a long and complex process as metrics are calculated at pre-defined time intervals of differing width and stride on patients with potentially week-long recordings. Additionally, patients' annotated electrocardiogram recordings require careful preprocessing and filtering steps to eliminate noise and erroneous values. Following an elaborate description of the utilized HRV method, we present an efficient C++ implementation. To analyze the execution time and efficiency of the implementation, a comparison with an existing HRV implementation in Python follows, to analyze the execution times for both solutions on a set of preprocessed patient databases. We present our results with a discussion of the benefits and drawbacks, concluding that the new implementation performs exceptionally better, with an average speedup of 198.7 and a median speedup of 81.4.
|
D. Mileski, N. Petrovski, M. Gusev (Sts Cyril and Methodius University in Skopje, Faculty of Computer Science and Engineering, Skopje, Macedonia) From Single to Multi-Node Multi-GPU Training: Performance Evaluation of HPC Distributed Training for ECG-based LLMs 
This study explores the transition from single-node, single-GPU configurations to multi-node, multi-GPU environments for training ECG-based large language models (LLMs) using high-performance computing (HPC) resources. It provides a detailed mapping of current frameworks for distributed deep learning in multi-node, multi-GPU settings, including Horovod from Uber, DeepSpeed from Microsoft, and the built-in distributed capabilities of PyTorch and TensorFlow. Performance is evaluated by comparing PyTorch's Distributed Data Parallel (DDP) approach with non-Distributed Deep Learning setups, focusing on scalability, speedup, efficiency, and resource utilization. The analysis leverages HPC infrastructure with SLURM, Apptainer (Singularity) containers, CUDA, PyTorch, and shell scripts to support training workflows and automation.
|
D. Miljković (Hrvatska elektroprivreda d.d., Zagreb, Croatia) Intelligence Beyond Human Limits: The Evolution and Implications of Machine Cognition 
This paper examines the evolutionary pathway from current narrow artificial intelligence (AI) systems through artificial general intelligence (AGI) to artificial superintelligence (ASI), with particular focus on the technological singularity as a potential watershed moment in human history. The analysis covers key technical milestones required for this progression, including advances in machine learning architectures, knowledge representation, and cognitive architectures. Drawing from recent developments in neuroscience and deep learning, the research proposes a framework for understanding the qualitative differences between these three stages of artificial intelligence development. The paper addresses critical questions about intelligence, consciousness and the challenges of maintaining human agency in a post-singularity world. Novel perspectives on safety mechanisms and technical guidelines necessary for managing the transition between these developmental stages are presented. The analysis suggests that while the path to AGI and ASI remains uncertain, the acceleration of technological progress introduces fundamental questions about the nature of intelligence and consciousness. The paper concludes by exploring potential trajectories of recursive self-improvement in artificial systems and their theoretical implications for the future of machine cognition.
|
D. Miljković (Hrvatska elektroprivreda d.d., Zagreb, Croatia) LLM-Enhanced Aircraft Piston Engine Diagnostics with Expert Knowledge Integration 
This paper presents a novel approach to aircraft piston engine diagnostics utilizing Large Language Model (LLM) enhanced with domain-specific knowledge. The methodology combines traditional engine monitoring data analysis with advanced natural language processing capabilities to identify and diagnose engine anomalies. The system processes real-time engine parameters including Cylinder Head Temperatures (CHTs), Exhaust Gas Temperatures (EGTs), engine RPM, Fuel Flow (FF), and Turbine Inlet Temperature (TIT) from engine monitoring systems. The LLM's diagnostic capabilities are augmented through the integration of expert knowledge, implemented as a comprehensive set of IF-THEN rules and detailed textual descriptions of common engine fault patterns extracted from maintenance documentation. This hybrid approach enables the system to leverage both structured operational data and unstructured maintenance expertise to provide accurate fault diagnostics. The methodology shows particular promise in detecting subtle fault patterns that may be missed by conventional monitoring approaches, potentially enabling more proactive maintenance interventions and enhanced flight safety.
|
I. Vlašić, G. Malčić, M. Lučan, D. Malez (Zagreb University of Applied Sciences, Zagreb, Croatia) Application of Open Source Libraries and MQTT Network Protocol in Simatic S7 Automation Systems 
It is common in the field of industrial computer automation to use proprietary software platforms provided by manufacturers of automation equipment, as this ensures the efficiency and reliability of their systems. However, integrators of automation systems now have the opportunity to adopt open-source technologies, which offer more flexible and cost-effective solutions. Although such solutions may not entirely replace traditional systems—primarily due to requirements for transparency and maintainability—they introduce new paradigms within the framework of Industry 4.0.
The MQTT communication protocol, due to its simplicity and adaptability, has become widely used in IoT systems, enabling implementation on microcontroller-based actuators and sensors. Additionally, open-source tools and libraries, such as Snap7, facilitate data integration from manufacturer platforms, like Siemens' SIMATIC S7 systems, via the Industrial Ethernet interface.
This paper describes a developed application that enables seamless integration of MQTT data into the native data formats of the SIMATIC S7 automation system without the need for programming. The solution contributes to the easier adoption of open technologies in industrial systems.
|
J. Ćurić (Fakultet elektrotehnike i računarstva, Zagreb, Croatia) Implementation of a Cost-Effective Smart Socket for Power Monitoring and Remote Control 
An embedded system for a smart power outlet has been developed, designed to measure the electrical power of devices connected to it. Power is determined by measuring voltage and current: the voltage is measured via a transformer for simplicity of implementation, while a Hall effect-based sensor is used for current measurement, enabling precise and safe measurement of high currents. Both sensors are connected to a 12-bit ADC. Although the voltage and current signals are noisy and inherently imprecise, meticulous data processing allows for obtaining more accurate results. The system provides a power measurement range starting from 3 W with a resolution of 1 W, demonstrating that reliable measurements can be achieved even with cost-effective components and non-ideal measurement methods. The smart outlet is built on a custom printed circuit board optimized for safely handling a maximum power of 2.4 kW, equivalent to 10 A at 240 V. All components through which the current flows were selected to prevent overheating. Smart functionality is achieved through remote control and result monitoring. The outlet is compatible with ZigBee networks, integrating automatically and correctly displaying within the network.
|
D. Budimir (Fakultet prometnih znanosti, Zagreb, Croatia), I. Peko (Veleučilište "Nikola Tesla" u Gospiću, Gospić, Croatia), D. Žižić-Gušo (Dalekovod d.d., Zagreb, Croatia) Oblikovanje suvremenih prometnih sustava: Studija na primjeru flote vozila Dalekovod d.d. 
Oblikovanje suvremenih prometnih sustava zahtijeva analizu operativne učinkovitosti utemeljenu na stvarnim prometnim uvjetima, pri čemu ključnu ulogu imaju podaci iz prometnih tokova, poznati kao FCD-podaci (Floating Car Data). U ovom radu istražuje se flota vozila tvrtke Dalekovod d.d., s posebnim naglaskom na dinamičke promjene u prometu kroz vremenske i prostorne aspekte prometne mreže. Primjenom višekriterijske analize omogućeno je vrjednovanje i usporedba različitih scenarija upravljanja flotom.
Rezultati istraživanja donose model koji omogućava relativnu usporedbu varijantnih rješenja, s ciljem povećanja operativne učinkovitosti i smanjenja troškova. Predloženi model predstavlja vrijedan alat za unapređenje upravljanja prometnim resursima te optimizaciju temeljem dubljeg razumijevanja prometne dinamike u stvarnim uvjetima.
|
|
Basic information:
Chairs:
Vlado Sruk (Croatia), Dejan Škvorc (Croatia)
Program Committee:
Goran Delač (Croatia), Željko Hocenski (Croatia), Leonardo Jelenković (Croatia), Hrvoje Mlinarić (Croatia), Adrian Satja Kurdija (Croatia), Vlado Sruk (Croatia), Marin Šilić (Croatia), Dejan Škvorc (Croatia), Klemo Vladimir (Croatia)
Registration / Fees:
REGISTRATION / FEES
|
Price in EUR
|
EARLY BIRD
Up to 23 May 2025 |
REGULAR
From 24 May 2025 |
Members of MIPRO and IEEE |
270 |
297 |
Students (undergraduate and graduate), primary and secondary school teachers |
150 |
165 |
Others |
300 |
330 |
The student discount doesn't apply to PhD students.
NOTE FOR AUTHORS: In order to have your paper published, it is required that you pay at least one registration fee for each paper. Authors of 2 or more papers are entitled to a 10% discount.
Contact:
Vlado Sruk
University of Zagreb
Faculty of Electrical Engineering and Computing
Unska 3
HR-10000 Zagreb, Croatia
Phone: +385 1 612 9945
Fax: +385 1 612 9653
E-mail: vlado.sruk@fer.hr
Dejan Škvorc
University of Zagreb
Faculty of Electrical Engineering and Computing
Unska 3
HR-10000 Zagreb, Croatia
Phone: +385 1 612 9946
Fax: +385 1 612 9653
E-mail: dejan.skvorc@fer.hr
The best papers will get a special award.
Accepted papers will be published in the ISSN registered conference proceedings. Papers in English presented at the conference will be submitted for inclusion in the IEEE Xplore Digital Library.

Location:
Opatija is the leading seaside resort of the Eastern Adriatic and one of the most famous tourist destinations on the Mediterranean. With its aristocratic architecture and style, Opatija has been attracting artists, kings, politicians, scientists, sportsmen, as well as business people, bankers and managers for more than 180 years.
The tourist offer in Opatija includes a vast number of hotels, excellent restaurants, entertainment venues, art festivals, superb modern and classical music concerts, beaches and swimming pools – this city satisfies all wishes and demands.
Opatija, the Queen of the Adriatic, is also one of the most prominent congress cities in the Mediterranean, particularly important for its ICT conventions, one of which is MIPRO, which has been held in Opatija since 1979, and attracts more than a thousand participants from over forty countries. These conventions promote Opatija as one of the most desirable technological, business, educational and scientific centers in South-eastern Europe and the European Union in general.
For more details, please visit www.opatija.hr and visitopatija.com.
|
|
|
Currently there are no news |
|
|
|
|