Search the digital library catalog

Abstract: V današnjem času vse večjega tehnološkega napredka postaja umetna inteligenca vse bolj priljubljena komponenta poslovnih modelov. V magistrskem delu smo obravnavali povezovanje umetne inteligence v CRM rešitve. Ugotavljali smo, kakšen vpliv imajo pametne rešitve, povezane z umetno inteligenco na odnose s strankami in samo poslovanje. V teoretičnem delu so predstavljene CRM rešitve-kot orodja in tehnologije, ki izboljšajo in upravljajo z odnosi s strankami. Umetna inteligenca pa je opredeljena kot področje računalniške znanosti, ki se osredotoča na razvoj inteligentnih strojev, ki bodo sposobni sprejemanja racionalnih odločitev in razmišljanja po človeškem vzgledu. Predstavljena je njuna integracija, delovanje in prednosti, ki jih ta povezava prinaša na poslovanje podjetij. V empiričnem delu smo izvedli primerjalno analizo treh ponudnikov rešitev CRM, ki imajo vključeno umetno inteligenco - Salesforce, Hubspot in Microsoft Dynamics 365. Rešitve smo podrobneje primerjali glede na njihove funkcionalnosti, ceno, možnosti integracije z ostalimi aplikacijami, proces implementacije, načine poročanja in kako je v rešitvi uporabljena umetna inteligenca. Povzeli smo rezultate primerjalne analize in pripravili napotke za izbiro prave CRM rešitve glede na velikost podjetja. V zadnjem poglavju smo se dotaknili še posebej pomembne tematike varstva podatkov in etičnih vprašanj, ki se porajajo z vedno hitrejšim razvojem umetne inteligence.
Keywords: CRM rešitve, umetna inteligenca, upravljanje odnosov s strankami, Salesforce, Hubspot, Microsoft Dynamics 365
Published in DKUM: 19.06.2025; Views: 0; Downloads: 3
Full text (3,43 MB)

Abstract: With computational modeling of lyophilization in vials, the pressure coupling between the sublimation front and the drying chamber has traditionally been calculated using a simplified mass transfer resistance model in the form of a model, which takes into account the headspace and the stopper in a simplified way. In developing a 3D CFD-based digital twin of lyophilization in vials, a need arises for a mass flow rate-dependent vial headspace/stopper model, as it enables a more accurate calculation of the pressure conditions above the shelf as well as pressure conditions directly at the sublimation front, the latter directly affecting the sublimation mass transfer rate as well as the temperature inside the product, which is crucial for determining the risk of product collapse. The local pressure variations at a shelf level affect the heat transfer conditions due to heat conduction in the low pressure environment of the drying chamber. In the present work the development of a coupled multilevel vial lyophilization model for the freeze-drying of vials is reported, with the time-dependent 1D heat and mass transfer model at the vial level coupled with the time-dependent 3D low-pressure CFD model of the flow of the water vapor–air mixture in the drying chamber heated by the shelves. A direct pressure coupling between the sublimation front and the drying chamber space in form of vial type specific headspace/stopper resistance model is implemented. The developed multilevel lyophilization model is used to study the pressure build-up above the shelf and the headspace of the vial and its influence on the product temperature at the bottom of the vial using simulations carried out for different chamber pressures (6 Pa and 22 Pa), shelf temperatures (−20 oC and +10 oC) and vial types (10R and 15R). By implementing previously developed vial headspace/stopper pressure resistance models, the computational results show that the pressure build-up above the shelf and vial headspace significantly affect the product temperature at the bottom of the vial, especially at low chamber pressures ( Pa) and small gap sizes between the rubber stopper and the shelf above it. The increased pressure outside the vial leads also to higher heat transfer by conduction, which is particularly pronounced at the central shelf positions and within smaller shelf gaps. These results underline the importance of using a coupled multilevel model when analyzing the relationship between the local pressure variations above the shelf and their direct influence on product drying conditions, further improving the predictive capabilities of CFD based multilevel lyophilization models, especially with respect to detecting the product collapse temperature.
Keywords: freeze-drying, conjugate heat and mass transfer, computational fluid dynamics, multi-scale modeling
Published in DKUM: 17.06.2025; Views: 0; Downloads: 4
Full text (4,01 MB)
This document has many files! More...

Abstract: Coarse-grained molecular dynamics simulations are highly valuable for studying protein-protein interactions. Unfortunately, commonly used force fields often overestimate these interactions. Here, we investigate the performance of the Martini 3 force field in predicting the self-interaction behavior of lysozyme and subtilisin using Metadynamics. The original Martini 3, despite improvements over its predecessor, overestimates interaction strength. Through reparameterization of bead interactions, we achieve good agreement with experimental data of the second virial coefficient and the diffusion coefficient. The new, refined force field enables more accurate CG-MD simulations, with potential applications in understanding and prediction of protein stability, aggregation tendencies, and solubility, with the possibility to aid in the development of protein-based drugs.
Keywords: coarse-grained molecular dynamics, Martini 3 force field, protein-protein interactions, B22, reparameterization, NMR, diffusion coefficient
Published in DKUM: 20.03.2025; Views: 0; Downloads: 2
Full text (6,83 MB)

Abstract: This study presents the synthesis and characterization of a series of multiblock copolymers, poly(ethylene 2,5-furandicarboxylate)-poly(ε-caprolactone) (PEF-PCL), created through a combination of the two-step melt polycondensation method and ring opening polymerization, as sustainable alternatives to fossil-based plastics. The structural confirmation of these block copolymers was achieved through Attenuated Total Reflectance Fourier Transform Infrared Spectroscopy (ATR-FTIR), ensuring the successful integration of PEF and PCL segments. X-ray Photoelectron Spectroscopy (XPS) was employed for chemical bonding and quantitative analysis, providing insights into the distribution and compatibility of the copolymer components. Differential Scanning Calorimetry (DSC) analysis revealed a single glass transition temperature (Tg), indicating the effective plasticizing effect of PCL on PEF, which enhances the flexibility of the copolymers. X-ray Diffraction (XRD) studies highlight the complex relationship between PCL content and crystallization in PEF-PCL block copolymers, emphasizing the need to balance crystallinity and mechanical properties for optimal material performance. Broadband Dielectric Spectroscopy (BDS) confirmed excellent distribution of PEF-PCL without phase separation, which is vital for maintaining consistent material properties. Mechanical properties were evaluated using Nanoindentation testing, demonstrating the potential of these copolymers as flexible packaging materials due to their enhanced mechanical strength and flexibility. The study concludes that PEF-PCL block copolymers are promising candidates for sustainable packaging solutions, combining environmental benefits with desirable material properties.
Keywords: poly(ethylene furanoate), poly(ε-caprolactone), block copolymers, thermal properties, molecular dynamics, crystallinity, mechanical properties, flexible packaging
Published in DKUM: 13.03.2025; Views: 0; Downloads: 6
Full text (4,73 MB)
This document has many files! More...

Abstract: When individual oscillators age and become inactive, the collective dynamics of coupled oscillators is often affected as well. Depending on the fraction of inactive oscillators or cascading failures that percolate from crucial information exchange points, the critical shift toward macroscopic inactivity in coupled oscillator networks is known as the aging transition. Here, we study this phenomenon in two overlayed square lattices that together constitute a multilayer network, whereby one layer is populated with slow Poincaré oscillators and the other with fast Rulkov neurons. Moreover, in this multimodal setup, the excitability of fast oscillators is influenced by the phase of slow oscillators that are gradually inactivated toward the aging transition in the fast layer. Through extensive numerical simulations, we find that the progressive inactivation of oscillators in the slow layer nontrivially affects the collective oscillatory activity and the aging transitions in the fast layer. Most counterintuitively, we show that it is possible for the intensity of oscillatory activity in the fast layer to progressively increase to up to 100%, even when up to 60% of units in the slow oscillatory layer are inactivated. We explain our results with a numerical analysis of collective behavior in individual layers, and we discuss their implications for biological systems.
Keywords: collective dynamics, coupled oscillators, dynamics of networks, network resilience, robustness, synchronization transition
Published in DKUM: 28.02.2025; Views: 0; Downloads: 428
Full text (5,87 MB)
This document has many files! More...