Institute of Informatics
AI-Driven Self-awareness and Cognition for Compute Continuum (AI4CC)
Adaptívna a reflexívna umelá inteligencia pre výpočtové kontinuum
| Duration: |
1.7.2024 - 30.6.2027 |
| Program: |
SRDA |
| Project leader: |
doc. Ing. Hluchý Ladislav CSc. |
| Annotation: | AI4CC aims to contribute with new AI-based methods and algorithms to the development of modern computing
continuums, addressing some of the key challenges of this domain. These challenges include enhancing the
autonomy and self-adaptation capabilities of compute continuum platforms to optimize decision-making and
responsiveness; adapting to changing conditions, necessitating flexible and adaptive computing models for
dynamicity; efficiently managing vast amounts of data, considering factors like decentralization, scalability, and
real-time processing for effective Data Management; ensuring transparency, explainability, and accountability in AI
and machine learning models within compute continuums; achieving seamless interoperability between diverse
devices and platforms within the continuum, impacting data exchange and communication for Interoperability, and
dealing with resource heterogeneity—variability in computing resources across the continuum, including edge and
cloud environments, posing challenges in optimizing performance and resource utilization. By systematically
addressing these challenges and pioneering advancements in these key areas, AI4CC aims not only to bridge
existing gaps but to propel compute continuums into a new era of efficiency, adaptability and security. The project
envisions a future where compute continuums seamlessly integrate into various domains, fostering a holistic and
intelligent computing environment that adapts to the ever-evolving demands of the digital landscape. |
CFD simulations of tunnel fire and wildland fire using FDS
CFD simulácie požiaru v tuneli a lesných požiaroch pomocou FDS
Digital Technologies for Critical Infrastructures (DICRIS)
Digitálne technológie pre kritické infraštruktúry
| Duration: |
1.8.2024 - 31.3.2026 |
| Program: |
|
| Project leader: |
doc. Ing. Hluchý Ladislav CSc. |
| Annotation: | To advance foundational scientific knowledge and design methodologies necessary for
accelerating the digital transformation of the Slovak power industry. This comprehensive research and
development effort aims primarily at creating and integrating digital twin technology for critical
infrastructure components of the Slovak electric transmission system. Additionally, it includes the
development of predictive analytical models and a context-aware decision support empowered with human
voice processing capabilities in Slovak. The ultimate goal is to enhance the safety, security, and reliability of
the electric power grid while driving broader digital transformation across various economic sectors. |
Dynamic malware analysis with explainable AI
Dynamická Malvérová Analýza s vysvetliteľnou AI
Environmental sensors based on 2D nanomaterials
Environmentálne senzory na báze 2D nanomateriálov
| Duration: |
1.8.2024 - 31.7.2026 |
| Program: |
SRDA |
| Project leader: |
RNDr. Kostič Ivan |
| Annotation: | The objective of this project proposal is the research of new semiconducting 2D materials for application in environmental sensors. 2-dimensional (2D) materials have been at the forefront of materials research in recent years due to their unic electrical and optical properties and interesting mechanical properties deriving from their atomically thin dimensions. One of promising applications of 2D materials are chemical, environmental, and biological sensor devices based on such 2D materials. In this project, we will focus on the development of environmental sensors based on 2D materials with concentration on gas sensors. |
Experimental investigation and computer modelling of airflow during road tunnel fire
Experimentálne skúmanie a počítačové modelovanie prúdenia vzduchu pri požiari v cestnom tuneli
| Duration: |
1.9.2025 - 31.8.2029 |
| Program: |
SRDA |
| Project leader: |
Mgr. Weisenpacher Peter PhD. |
| Annotation: | The research project will contribute to the verification of the computer simulation of the airflow and smoke spread in road tunnels and to the increase of fire safety of tunnels in Slovakia. The project aim is to investigate the interaction of emergency ventilation operation, meteorological factors and fire in a real tunnel and to create a series of computer simulations of customer-designed fire scenarios in two motorway tunnels, which will be verified by in situ measurements. Detailed studies on the course of fires and the efficiency of emergency ventilation will be developed. The simulations will be carried out on the HPC infrastructure at the Slovak Academy of Sciences. In order to validate the results of the created simulations, full-scale experiments and measurements of airflow and meteorological factors will be carried out by the customer of the project results. Modifications to the Tunnel Traffic & Control Simulator (TTOS) at the University of Zilina in Zilina will be designed and implemented in accordance with current legislation, which will strengthen its uniqueness in the Central European area and improve its use for training and testing tunnel control operators in Slovakia. Part of the solution will be the recording and evaluation of emergency events in tunnels in Slovakia and the development of a series of visualizations of the course of fires based on computer simulations for TTOS, which will serve as a visualization didactic tool in the education of tunnel control operators and university students. The project solution will make it possible to intensify and streamline the research cooperation of the project solvers, to create a strong partnership of research organizations with important customers of the project results, and to effectively use the modern research infrastructure of the participating organizations to create innovative solutions to increase the safety of road tunnels. |
Experimental System for Wireless Stimulation and Monitoring of Selected Biological Properties and Cognitive Abilities of Drosophila melanogaster.
Experimentálny systém pre bezkontaktnú stimuláciu a monitorovanie vybraných biologických vlastností a kognitívnych schopností Drosophila melanogaster.
| Duration: |
1.9.2024 - 30.6.2027 |
| Program: |
SRDA |
| Project leader: |
Ing. Mgr. Andok Robert PhD. |
| Annotation: | The fruit fly Drosophila melanogaster serves as one of the most versatile models for studying human diseases, including neurodegenerative or learning disorders. Thanks to the unique genetic tools available exclusively in the fly model, Drosophila considerably contributed to discoveries of genetic regulations behind processes such as learning and behavior. Nevertheless, progress in this research area is limited by available technologies for the stimulation and tracking fruit flies. This project aims to develop novel equipment for wireless stimulation and monitoring of insect behavior. A wireless micro-heater will be installed onto the Drosophila’s body, designed and developed within this project, together with the microsensor of position / force. The fly will be stimulated individually under certain circumstances by locally controlled heat induced in the micro-heater realized by connecting a nanometric diamond film (as a supporting material bilogically compatible with the examined tiny insects, and at the same time a material with excellent thermal conductivity) with a high density micro coil consisting of a closed LC resonant circuit with a high-k dielectric material (ZrO2, SrTiO3, HfO2) used for the micro-capacitor. By exerting electromagnetic waves of certain resonant frequency related to the size of the coil, electromagnetic induction will occur and the device will heat up stimulating the insect. The project aims to optimize and utilize this technology for the study of olfactory and social learning. The technology will nonetheless have a broad impact in other areas of Drosophila neuroscience, ethology and physiology. The system can be, for example, used also for wireless stimulation of flies in various types of learning experiments, for tracking fly behavior in complex environments, or for studies of behavioral roles of diverse genes and environmental factors. The suggested technology will have broad implications in both primary research and biomedicine. |
Smart Data PIpelines for the Cognitive ComputE Continuum (SPICE)
Inteligentné dátové kanály pre kognitívne výpočtové kontinuum
| Duration: |
1.4.2025 - 30.9.2027 |
| Program: |
|
| Project leader: |
doc. Ing. Hluchý Ladislav CSc. |
| Annotation: | The growing scale and complexity of the Cloud led to the emergence of Edge and IoT computing in the "compute
continuum" for applications relying heavily on sensor data and time-critical AI-assisted data processing. Its primary
goal is to enable intelligent decision making and gain deeper insights across the continuum. At the same time,
assembling and maintaining well-designed data pipelines involves addressing numerous technical, operational and
organizational challenges. These hurdles and expertise/resource constraints act as a barrier for wider adoption of data
pipelines. Our goal is to provide: (1) AI-assisted composition of smart data pipelines to radically simplify complex
and distributed data processing, (2) a highly optimized execution environment and resource management that enable
secure and smart data processing for the compute continuum. The results of our project will be made available as an
innovation transformation incubator for others to exploit. |
Intelligent sensor systems and data processing
Inteligentné senzorové systémy a spracovanie dát
| Duration: |
1.1.2023 - 31.12.2026 |
| Program: |
VEGA |
| Project leader: |
Ing. Malík Peter PhD. |
| Annotation: | The central theme of Industry 4.0 and 5.0 is the digitization, intelligence and decentralization of management, so a key research is the new generation of smart sensors, able to cooperate and adapt to environment changes. This will be achieved by researching new methods of aggregating hyperspectral and multimodal data, as well as algorithms using artificial intelligence. The project is focused on intelligent algorithms for non-contact surface sensing in high-noise environments, which are able to learn the nature and noise distribution from data. This results in higher accuracy and greater noise robustness. The emphasis is on the classification and anomaly detection, which will bring more accurate and robust algorithms for use with the high noise content and long-tailed distribution that dominates in the common industrial environment. Research into aggregation algorithms for heterogeneous and multisensor data will bring new compensation mechanisms to suppress the effects of negative factors on sensor systems. |
Computer simulation of airflows and fire smoke spread in critical structures
Počítačová simulácia prúdenia vzduchu a šírenia dymu pri požiari v kritických objektoch
| Duration: |
1.1.2024 - 31.12.2027 |
| Program: |
VEGA |
| Project leader: |
Mgr. Weisenpacher Peter PhD. |
| Annotation: | Research in the proposed project is focused on formulation of new scientific knowledge on computer simulation of
airflows and fire smoke spread in critical structures. Motorway tunnels were selected as the main subject of
research interest based on discussions with specialists on fire safety in Slovakia. Tunnels belong to inteligent
structures with high safety requirements due to potentially huge losses in case of fire. Natural airflows, airflows
created by emergency ventilation, airflows induced by fire and fire smoke spread will be analyzed with focus on
velocity fields, velocity profiles and smoke stratification. Computational aspects of efficient parallel realization of
computer simulation on HPC systems will be investigated as well. The previous research results, experience and
obtained experimental data will be utilized. The research is in line with current research trends and requirements
of the fire researchers’ and simulators developers’ community and has potential to have significant social impact. |
Progressive methods of the transfer of nanostructured semiconductive 2D materials based on transition metal dichalcogenides onto microelectronic elements
Progresívne metódy transferu nanoštruktúrnych polovodivých 2D materiálov na báze dichalkogenidov tranzitných kovov do mikroelektronických prvkov
| Duration: |
1.1.2022 - 31.12.2025 |
| Program: |
VEGA |
| Project leader: |
Ing. Mgr. Andok Robert PhD. |
| Annotation: | The aim of this project is to carry out basic research in the field of new progressive nanostruct. semiconductive materials based on dichalcogenides of transition metals with the focus on nanostructured disulfides. The properties of selected nanostructured disulfides will be examined in terms of their use in microlelectronics and expected advantages of nanostructured disulfides in comparison with bulk semiconductor materials will be shown.
We will design model microelectronic devices based on specific nanostructured disulfides such as WS2, MoS2, MSe, and develop technological methods for their preparation. We will master mechanical exfoliation of nanostructured disulfide layers and transfer of these nanostructured layers to the microelectronic device on the substrate. We will also focus on the analysis of these layers and their structural properties by physical methods (SEM, AFM, EDX, Raman spectroscopy...) and on the characterization of electrical and transport properties of the model microel. structure. |
Semantic distributed computing continuum for extreme data processing
Sémantické distribuované výpočtové kontinuum pre spracovanie extrémnych dát
Structures S3PR, ES3PR and S4PR of mathematical models of resource allocation in discrete production systems based on Petri nets and their use for deadlock prevention.
Štruktúry S3PR, ES3PR a S4PR matematických modelov alokácie zdrojov v diskrétnych výrobných systémoch na báze Petriho sietí a ich využitie na prevenciu deadlokov.
| Duration: |
1.1.2025 - 31.12.2028 |
| Program: |
VEGA |
| Project leader: |
doc. Ing. Čapkovič František CSc. |
| Annotation: | Discrete systems remain in a certain state until they are forced to change this state due to the occurrence of some discrete event. They are called DES (Discrete-Event Systems). They include a wide class of real systems – flexible (alias automated) manufacturing systems, robotic cells, communication systems, transport systems, etc. Petri nets PN are very suitable for mathematical modelling of DES. RAS (Resource Allocation Systems) in DES tend to deadlocks. PN-based DES models can be used to eliminate deadlocks. Three model paradigms, namely S3PR, ES3PR and S4PR, will be investigated using two approaches: (i) structural analysis of PN RAS models, specifically using siphons and traps, without the need to know the initial state, (ii) P-invariants PN models of RAS and reachability tree (RT) of states. Both approaches will be applied to all model paradigms, compared and evaluated using simulations in Matlab (or SciLAB). |
(SILVANUS-SK)
Výskum a vývoj vybraných inovatívnych technológií pre platformu manažmentu lesných požiarov
Stopping criteria to bound distributed consensus algorithms with asymptotic convergence for network size estimation
Zastavovacie kritériá pre ohraničenie distribuovaných konsenzuálnych algoritmov s asymptotickou konvergenciou pre odhad veľkosti siete
| Duration: |
1.4.2024 - 31.12.2025 |
| Program: |
SRDA |
| Project leader: |
Ing. Kenyeres Martin PhD. |
| Annotation: | Knowing the network size (or at least its precise estimate) beforehand is crucial for many modern distributed
algorithms executed in multi-agent systems. As seen in the literature, widespread consensus-based algorithms for
distributed averaging can be easily applicable for this purpose. However, they have not been too frequently used to
estimate the network size since their definition. The research of the researchers involved in this project is planned
to be focused on how to efficiently stop the execution of these algorithms for network size estimation in a
distributed way. The efficient operation of algorithms is one of the most crucial design requirements placed on
topical multi-agent systems, as identified in many related manuscripts. The primary goals of this project are to
analyze the mentioned algorithms for network size estimation from numerous aspects (e.g., estimation precision,
convergence rate, robustness, etc.), examine and optimize the performance of the existing stopping criteria, and to
propose novel stopping criteria that optimize consensus-based data aggregation in multi-agent systems. Multiagent systems in this project will be modeled as graphs with random topologies, ensuring a credible representation
of real-world systems. An analysis of the used methods' reliability using probabilistic tools such as large deviations
is also planned to be done. |
Getting the right info on ticks (INFOTICK)
Získanie pravdivých informácií o kliešťoch
| Duration: |
1.7.2023 - 30.6.2027 |
| Program: |
SRDA |
| Project leader: |
Ing. Gatial Emil PhD. |
| Annotation: | Despite the fact that the castor bean tick, Ixodes ricinus has been studied for a century, many questions regarding
its ecology remains unanswered. Several aspects of its basic biology and phenology are still unexplored. Global
changes, including climate shifts, transformation of the landscape and urbanization, contribute to the switch not
only in tick distribution, but also in bionomics and seasonal activity of ticks. The ornate dog tick, Dermacentor
reticulatus adapts quickly to changing conditions and its range is expanding. There is the need for detailed
description of areas where these ticks are found (natural as well as urban habitats), since their ranges have
changed during the last decades. The main risk factor for tick -exposed people in a given area is the density of
infected questing ticks. In the proposed project, questing activity of ticks will be monitored using the tick -plot
methodology „tick gardens“ in field plots as well as flagging the vegetation for questing ticks. Using the tick-plot
methodology, we will also follow the tick life cycle and the seasonality of various developmental events (especially
moulting) as well as the longevity of different life stages. Since these two species of ticks are considered
epidemiologically the most important, we will also identify the prevalence and occurrence of both pathogen infected
questing ticks and infected ticks feeding on animals. Furthermore, with changing conditions, the invasion and
occurrence of „non-native“ species of ticks in Slovakia will be closely monitored since these emerging tick species
can introduce new pathogens to our area. The information obtained by the research team during the project as well
as during previous studies will be transferred and used in the development of a mobile application for tick
identification and the creation of a website that will bring benefits to the general public and professionals to
understand the risk of infection with the tick-borne pathogens. |
The total number of projects: 16
