73595 research outputs found
Sort by
Application of the AI in Logistics
No full textTato bakalářská práce se zaměřuje na aplikaci umělé inteligence v oblasti logistiky, konkrétně
v prostředí malého podniku VALID s.r.o. Cílem práce bylo analyzovat aktuální logistické pro-
cesy, identifikovat jejich slabá místa a následně navrhnout efektivní řešení, využívající tech-
nologie AI a IoT. Práce se věnuje jak popisu současného stavu integrace AI v logistice, tak i
návrhu autonomních systémů řízení zásob, včetně porovnání tradičních a moderních přístupů k
modelování zásob. Na základě provozních dat byl navržen koncept chytrého skladu, s cílem
zvýšit efektivitu, snížit chybovost a zrychlit práci skladníků. Výsledný návrh obsahuje i
ekonomické vyhodnocení, včetně výpočtu návratnosti investice a technický postup implemen-
tace. Součástí práce je rovněž testování různých chatbotů při predikci zásob a vytvoření riziko-
vého rámce pro používání AI ve firmě.This bachelor's thesis focuses on the application of artificial intelligence in the field of logistics,
specifically within the environment of the small company VALID s.r.o. The objective of the
thesis was to analyse the current logistics processes, identify their weak points, and subse-
quently propose an effective solution utilizing AI and IoT technologies. The thesis covers both
the current state of AI integration in logistics and the design of autonomous inventory manage-
ment systems, including a comparison between traditional and modern inventory modelling
approaches. Based on operational data, a smart warehouse concept was designed with the goal
of increasing efficiency, reducing errors, and accelerating warehouse operations. The resulting
proposal also includes an economic evaluation, including the calculation of return on invest-
ment and a technical implementation procedure. The thesis further includes testing of various
chatbots for inventory prediction and the creation of a risk framework for the safe application
of AI within the company
Fearless Steps Challenge: Speech Detection in NASA Apollo Program Audio Recordings
No full textTato bakalářská práce se zabývá úlohou detekce řečové aktivity, tedy identifikací časových úseků ve zvukovém signálu, během nichž je přítomna mluvená řeč. Cílem je navrhnout model hlubokého učení pro detekci řeči v reálných datech z misí Apollo poskytnutých ve čtvrté fázi výzvy Fearless Steps Challenge. Práce popisuje proces načtení zvukových dat, převod signálu na MFCC příznaky, rozdělení příznaků do překrývajících se segmentů a další zpracování a přípravu dat pro detekci řečové aktivity pomocí modelů hlubokého učení. Navržené modely jsou trénovány, validovány a testovány na poskytnutých datových sadách. Pro dosažení co nejlepších výsledků bylo porovnáno několik různých architektur, přičemž nejvyšší výkonnosti dosáhl model architektury Conformer. Tento model je následně porovnán s publikovanými výsledky v rámci uvedené výzvy. Přestože trénování bylo ztíženo absencí evaluační datové sady, úspěšnost navrženého modelu se přiblížila k nejlepším publikovaným přístupům.This bachelor's thesis addresses the task of voice activity detection, which involves identifying time segments within an audio signal during which spoken speech is present. The goal is to design a deep learning model for detecting speech in real-world data from the Apollo missions, provided as part of Phase 4 of the Fearless Steps Challenge. The thesis describes the process of loading the audio data, converting the signal into MFCC features, dividing the features into overlapping segments, and further processing and preparing the data for voice activity detection using deep learning models. The proposed models are trained, validated, and tested on the provided datasets. To achieve the best possible results, several different architectures were compared, with the Conformer architecture achieving the highest performance. This model is then compared with published results from the challenge. Despite the lack of an evaluation dataset complicating the training process, the performance of the proposed model approached that of the best published approaches
Preparation and characterization of magnetron sputtered layers on polymer substrates
No full textTato diplomová práce se zaměřuje na přípravu a hodnocení tenkých vrstev nanesených metodou magnetronového naprašování na skleněné a polymerní substráty. Cílem bylo optimalizovat složení vrstev za účelem zlepšení jejich tribologických a antibakteriálních vlastností, zejména pro aplikace ve zdravotnictví. Výzkum probíhal ve dvou fázích: nejprve byly na skleněné substráty naneseny základní vrstvy obsahující titan, měď a stříbro, následně byly vytvořeny jejich oxidické varianty pomocí reaktivního naprašování na substráty z polyamidu. Nejlepších výsledků bylo dosaženo u vrstev s vyšším obsahem kyslíku a vyváženým poměrem stříbra a mědi, které vykazovaly nízký součinitel tření a zároveň silnou antibakteriální aktivitu. Výsledky potvrzují, že oxidické vrstvy připravené za vhodných podmínek depozice mají vlastnosti, které je činí perspektivními pro využití v biomedicínských aplikacích.This thesis focuses on the preparation and evaluation of thin films deposited by magnetron sputtering onto glass and polymer substrates. The aim was to optimize the composition of the films to improve their tribological and antibacterial properties, particularly for medical applications. The study was conducted in two stages: first, base layers containing titanium, copper, and silver were deposited on glass substrates, second, their oxidized variants were
prepared via reactive sputtering on polyamide substrates. The best results were obtained for coatings with higher oxygen content and a balanced composition of silver and copper, which exhibited a low coefficient of friction combined with strong antibacterial activity. The results confirm that oxidized coatings prepared under suitable deposition conditions possess properties that make them promising for use in biomedical applications
Educationalists in 1950s, 1960s Hungary: Identity and Profession through Retrospective Life (Hi)Stories
Get PDFThis paper is based on the lost and found sources from the former Hungarian National Pedagogic Institute (OPI) and the National Educational Institute (OKI): due to the generosity of Professor Gábor Halász we can save valuable archives from disappearance. I found several life-story interviews in the corpus by different Hungarian educationalists, which were recorded in 1984 and 1985. These documents gives a unique opportunity to describe typical career patterns, how a teacher could become a head of a department in the Ministry or held other key positions on the highest level of educational administration in the post-war, communist Hungary. The semi-structured interviews focused on personal transitions and turns, connected with the historical time and political changes, including different socio-historical contexts and dimensions, like the recalled decades of 1950s and 1960, the interviewing period of the late socialism, and finally, the retrospection of the questioners. The biographic nature of the communist political system is a specific characteristics in my analysis: to get a position (like an editor in chief, a school inspector of different districts in Budapest, or a head of a department in the Cultural Ministry) it was required to constantly write and rewrite autobiographies; construct social and professional identities again and again, proved loyalty to the Party. The contemporary reader from 2023 can evaluate the sources as narrations: in the beginning crisis of the system in the 1980’s, the interviewees told their lives with the intention to form the memory, create legitimation and meanings for their past activities and life-stories
Manipulation of Time Continuity in Shared Narratives. On the Construction of Collective “Truths” and Its Ambivalent Function in the Social World and in Education
Get PDFThis article discusses the topic of politically motivated manipulation of time continuity in shared narratives. It examines the consequences of the irrational fusion of fragments of the stories of different protagonists who live in different historical periods, sometimes separated by centuries. The integration of bits and pieces of the myths borrowed from the past into the narratives of the present can take on psychotic proportions. It can seriously damage the living tissue of personal and collective memory. The story of the past, told in a certain way, can influence the self-perception of people, making them feel partly like heroes with a special mission, as well as victims, threatened by a hostile world, isolated, acting in response to the past, and therefore out of touch with reality. The result is the impossibility of distinguishing current events as a set of specific political and psychological forces that require a specific response to an urgent situation (Kalinowska, 2012). Current threats are thus perceived not only in their specificity, but in constant vigilance in the light of past traumas and illogical linking of the past, present, and future. The goal of this article is the understanding of this process with the help of Vamik Volkan’s psychoanalytic concept of “time collapse” (Volkan & Javakhishvili, 2022), and Michael Rothberg’s theory of the “implicated subject” (2019). When we talk about the transition to democracy and solidarity in education, we must understand the way of thinking that deviates from them
Navigating urban logistics challenges: An optimized approach to parcel distribution in the Prague city center
Get PDFIn this paper, we focus on metropolitan transport logistics that has always been difficult due to the need to transport goods through a complex urban infrastructure, heavy traffic, and dense populations. The goal of this paper is to describe a micro hub location in the neighborhood of the historic center of the city of Prague, the Czech Republic in a situation where the municipalities consider prohibiting the entry of trucks with combustion engines that currently provide cargo transportation for B2B partners operating in this area. Micro hubs represent the efficient last mile consolidation and distribution facilities located in or near urban neighborhoods and serving a spatially limited, densely populated delivery area. We propose a solution combining facility location problem, vehicle routing problem and balancing with the horizontal and vertical cooperation in supply chains to locate the micro hub in the area with extremely low availability of suitable space while respecting the necessity to connect this facility to an existing network of regional terminals operated by 3PLs to ensure the distribution of cargo across the board according to the requirements of customers. For the collection and distribution of parcels within the area, we suggest to use of cargo bikes and electric vans to provide environmentally sustainable service and minimize the harmful consequences of excessive traffic for residents. We also discuss the economic implications of adopting such innovative sustainable supply chain solutions for involved horizontal and vertical supply chain partners emphasizing different motivation aspects to convince these partners to cooperate and share scarce resources
DuPont analysis among European dentistry companies to measure the impact of the COVID-19 pandemic
Get PDFAlthough health economics belongs to the highly respected economic disciplines within the research literature, there is a considerable gap related to the investigation of dentistry in particular, even after the global pandemic of COVID-19 disease. Fundamentally, the DuPont framework is a well-known complex analysis to evaluate companies from the point of view of financial performance. The investigation of the return on equity as a relation between the return on assets and the equity multiplier, simply called the leverage effect, is presented in this paper. Therefore, this study aims to estimate the effect of leverage and its changes due to the COVID-19 pandemic among dentistry companies in selected European countries with different healthcare insurance systems. This comparative investigation focused on the generalised method of moments with dynamic panel data from Orbis, the Bureau van Dijk financial database for 1,128 dentistry companies in nine European countries. Methodologically, concerning those post-estimation techniques to evaluate over- and underestimation of the models. It has been differentiating between companies with a high or low ownership concentration structure. The results have shown differences in the leverage effect during the pandemic, assuming that companies with a major owner increased their equity, while the debt leverage increased among those companies with dispersed ownership and vice versa. If economic theory states that debt financing is more effective for a company than using internal sources, it is apparently different in the case of dentistry during the pandemic. However, dispersed ownership is more often related to dentistry, according to mergers in this particular business industry
WOUND DRESSING WITH TEXTILE DRESSING APPROACH: A REVIEW
Get PDFTo help healing, protect and care the wound from additional damage, a sterile dressing is applied to wound and injured area. Wound dressing can be a sterile pad or compress that directly contact with the wound area. Nowadays, different models of wound dressings are used in the medical field. In this review paper different wound dressings and efforts made on the textile based wound dressings discussed