In-sensor computing as a new paradigm for edge electronics


10:00 am - 11:30 am

Event Type

Pozývame Vás na


Ústavu informatiky SAV, v.v.i. ktorý sa bude konať v stredu 10. 5. 2023 o 10.00 hod. v zasadačke č. 102


Ing. Boris Hudec, PhD. (Elektrotechnický ústav SAV, v.v.i.)

Insensor computing as a new paradigm for edge electronics

Inspired by nature, the concept of insensor computing combines traditionally distinct acts of sensing and signal processing into one operation, in a similar fashion in which inmemory computing combines storing and information processing. This is achieved by employing multiple sensing devices with engineered responsivities as synaptic connections in a simple neural network. Engineered responsivity of each sensor, and their arrangement within a synaptic matrix, allow to hardcode simple processing algorithms into the sensing network. This approach would allow for edge sensing devices to output signals which are already preprocessed to a certain degree, greatly reducing the requirements for processing and upstream bandwidth. In this talk, we will discuss how simple yet revolutionairy insensor computing systems can be built bottomup utilizing local technology and talent.

Tešíme sa na stretnutie s Vami pri šálke kávy alebo čaju.

Linka na on-line pripojenie:–qmka–ahj

Ing. Mgr. Robert Andok, PhD.

SEMINÁR Ústavu informatiky SAV, v. v. i.


10:30 am - 12:30 pm

Event Type

Matej Babič, PhD.
Faculty of information studies, Novo mesto, Slovenia

New composite intelligent system of machine learning

One of the most important characteristics of the microstructure of a material is hardness. The analysis of structure is based on statistical methods and on the principle: from the general image of the structure to the particular (individual grains) by averaging. With computer methods, material structure is characterized by determination complexity of microstructure. The aim of this presentation is to present machine learning methods to predict mechanical  properties of robot laser hardened specimens with a new technology of hardening, a novel approach for pattern recognition by using fractal geometry. Microstructures were obtained from SEM images of robot laserhardened specimens from which we calculated the fractal dimension of 3D objects. I will present process of robot laser hardening under different angles of laser beam.

CVMatej Babič received his Ph.D. degree in Computer Science in 2014 from the Faculty of  Electrical Engineering and Computer Science of the University of Maribor, Slovenia. He studied Mathematics at the Faculty of Education in Maribor, Slovenia. He concluded postdoc at Jožef Stefan Institute, Ljubljana, Slovenia. At this time, he is Assistant Professor at Faculty of  information studies, Novo mesto, Slovenia. His research interest is in fractal geometry, graph theory, intelligent systems, hybrid machine learning, public transport and topography of  materials after hardening.

Location: online at

Seminár ÚI SAV


10:30 am - 12:00 pm

Event Type

Pozývame Vás na


ktorý sa bude konať v utorok 28. 2. 2023 o 10.30 hod. v zasadačke ÚI SAV č. 102.

1. Triboelectric Nanogenerator with Aligned Carbon Nanotube

RueiChi Hsu, Ph.D. Department of Power Mechanical Engineering,
National Tsing Hua University, Hsinchu, TAIWAN

2. Field Emission characteristics of laser selective growth CNFBCNT

YiHung Chen (Ph.D. Candidate) Department of Power Mechanical Engineering, National
Tsing Hua University, Hsinchu, TAIWAN

1. A triboelectric nanogenerator (TENG) is one of the most promising approaches for selfpowered energy conversion devices, which could transfer mechanical energy into electrical energy. Besides, this wellknown technique has been brought to attention concerning its highpowered output, low cost, sustainability, and wearable application. This wearable device could be attached to the human body to generate energy by walking, running, or any friction movement.

Every material has either triboelectric positive or triboelectric negative characteristics. When the difference between the friction affinity of two materials is enormous, the produced electrical charges will enhance the output performance. On the other hand, another improvement in output performance TENG can be attributed to the triboelectric layer’s surface contact area and the surface charge density.

We proposed to use thermal chemical vapor deposition (CVD) to grow a vertically aligned carbon nanotube (CNT) as a triboelectric layer. CNT has the advantage of good conductivity, flexibility, and mechanical property, which can be used as field emission devices, actuators, and sensors. One main reason that affects the growth of CNT is the buffer layer deposited on the silicon wafer. It is found that the Al2O3 buffer layer has a significant impact on the catalyst particle with a small average size, a large density, and a uniform size distribution. Hence, the annealing treatment on the Al2O3 buffer layer was investigated and will be discussed further.

KeywordsTENG, CNT, Al2O3 buffer layer, Thermal CVD

2. Field emission is one of the electronemitting methods that recently have been applied to the scientific or industrial fields, such as the electron microscope, microwave amplifier, display, cathode ray tube, plasma generator, and source of xray. When an electrical field is applied to the material, it creates the tunneling effect or Schottky effect on the electrons at the surface, increasing the possibility of these electrons conquering the energy barrier with energy lower than required and emitting to the environment. Since the field emission effect is highly related to the electrical field, proper emitter distribution guarantees better emission efficiency and reduces the screening effect. As a result, many previous studies are dedicated to improving the field emission properties by adjusting the density and distribution of the emitters or changing the material with lower work functions.

This study uses the carbon nanoflake ball (CNFB) as the emitter’s material. CNFB is a novel carbonbased composite material discovered by PingHuan Tsai, 2018. It is a spherelike material constructed with multidirection carbon composite flakes. Each flake has a diamond core covered by a few graphene layers. During the field emission process, the diamond core provides a robust construction with good heat conduction, while the graphene layer transports the electrons with lower resistance. Another unique characteristic of the CNFB is the synthesis mechanism. When synthesizing CNFBs with microwave plasma chemical vapor deposition (MPCVD), the nucleation sites of the CNFB are highly possible to be generated at the surface’s defect.

With the properties of the CNFB, this study proposed a method of using the laser to create designed defect patterns on the silicon surface to study the selective growth of CNFBs, and the relationship between the laser and the deposition of CNFB is discussed. To further improve the field emission characteristics of the CNFB, A layer of carbon nanotubes (CNT) is deposited on the patterned CNFBs. The result of the field emission shows a promising improvement in the turnon field voltage and the threshold voltage with the designed pattern.

Keywords: Field emission, Carbon nanoflake ball, Laser, MPCVD, CNT

Tešíme sa na stretnutie s Vami pri šálke kávy alebo čaju.
Ing. Mgr. Robert Andok, PhD.


Cestovateľský seminár


10:30 am - 12:00 pm

Event Type

Pozvánka na cestovateľský seminár

Milé kolegyne a milí kolegovia, pozývame Vás na zaujímavý cestovateľský seminár, ktorý sa bude konať v stredu 18.1. 2023 o 10.30 hod. v zasadačke ÚI SAV č. 102.

Haiti: Čo nevieme o tejto krajine.
Ing. Jean-Rosemond DORA, PhD. (odd. L.Hluchý)

  1. Mesto kde som vyrastal, možnosti vzdelávania, životné podmienky, zvyky, tradície
  2. historické pamiatky, pláže, príroda,
  3. video (4-5 minút)

Tešíme sa na príjemné posedenie a zaujímavé rozprávanie pri šálke kávy alebo čaju.

Seminár “Superpočítačové modelovanie materiálov”


10:00 am - 12:00 pm

Event Type

Pozývame Vás na SEMINÁR ÚI SAV,

ktorý sa bude konať v utorok 29. 11. 2022 o 10.00 hod. v zasadačke č. 102

Superpočítačové modelovanie materiálov

prof. Ing. Ivan Štich, DrSc., ÚI SAV

Seminár v skratke uvedie simulačné techniky, ktoré používame na superpočítačové modelovanie materiálov na atomárnej úrovni. Numericky najnáročnejšia časť modelovania je riešenie Schrödingerovej rovnice, ktorú riešime buď v jednočasticovom priblížení metódami teórie funkcionálu hustoty (DFT) alebo v mnohočasticovom priblížení stochastickými metódami kvantového Monte Carla (QMC). Metódy QMC predstavujú alternatívnu paradigmu riešenia mnohočasticového problému, ktorá kombinuje ultravysokú presnosť s ideálnou a neobmedzenou škálovateľnosťou s počtom CPU/GPU. Aplikáciu simulačných metód budeme demonštrovať výsledkami zo simulácií v oblasti elektrónovej štruktúry 2D materiálov a oblasti zobrazovania a atómárnej a nábojovej manipulácie povrchov a nanoštruktúr na nich vytvorených technikami bezkontaktnej atomárnej mikroskopie.

Prednáška bude v anglickom jazyku.

Prihlásenie je možné aj cez link:

Pozvánka v PDF