139585 research outputs found
Sort by
Bringing the entrepreneurial voice to the fore in entrepreneurial ecosystems
No full textPublisher Copyright: © James A. Cunningham, Matthias Menter, Conor O’Kane and Marco Romano 2024. All rights reserved.Recent entrepreneurial ecosystems literature has helped push general entrepreneurship research away from the idea of the entrepreneur as a lone Schumpeterian hero figure and advanced a view of entrepreneurship as a social process connected to a local context. However, that same literature often portrays entrepreneurs as objects-individuals who are acted upon through support from public-sector experts. In this study of a regional entrepreneurial ecosystem, we argue that the voice of the entrepreneur as an active subject is largely absent from existing literature. We argue that in terms of building entrepreneurial ecosystems, both practitioners and the research community should pay more attention to the diversity of the entrepreneurial voice. We identify five very different entrepreneurial voices that bring diversity to the dominant narrative of the entrepreneurial ecosystem as a melting pot for new technology and high-growth start-ups. In addition, we suggest further study of how entrepreneurial ecosystems acquire legitimacy in the eyes of entrepreneurs and how this legitimacy should be built, orchestrated or managed. Further, we discuss the implications of entrepreneurial voice for research, policy and practice related to entrepreneurial ecosystem building.Peer reviewe
Generative AI and information privacy : Users' assessment of the risks and benefits of information disclosure
No full textGenerative Artificial Intelligence (AI) has been valued for its interactive and advanced multimodal capabilities, and it has observed wide adaptation in recent years. Alongside this development, issues related to information privacy have been highlighted, as generative models can remember personal information from the training data and leak that information. Further information privacy issues arise for the users of generative AI from the interactive nature and advanced capabilities that can encourage users to share more information. This highlights important questions about how user input is collected and stored for purposes such as model development. The objective of this thesis is to study which factors affect users’ perception of the risks and benefits of information disclosure with generative AI.
A research model was developed based on the privacy calculus framework and extended with constructs from technology acceptance model. The data was collected with a survey and analysed with Partial Least Squares Structural Equation Modelling (PLS-SEM). Most of the hypotheses were supported and the results highlight that technical features, organizational practices and regulation can influence users’ intention to disclose information when interacting with generative AI. In more detail, high risk perceptions lowered the intention to disclose, while perceived benefits increased it. Expectations for regulation heightened risk perceptions, whereas trust in the AI provider’s data handling practices reduced the risk perception. Additionally, ease of use and social presence positively influenced the perceived benefits of information disclosure. Contrary to previous research, motor impulsivity did not have a significant effect on the intention to disclose.
This thesis contributes to the existing research by offering insights into the factors affecting information disclosure decisions with generative AI contexts by extending the privacy calculus model with AI-specific constructs. For practical implications, the results suggest that excessive risk may diminish users’ sense of privacy, which could make the user withhold from interacting with generative AI. On the other hand, increased perceived benefits of information disclosure may also lead to privacy and security risks. For organizations, clear guidelines and awareness about data use can support the adaptation and secure utilization of generative AI. For policymakers, it is important to support AI literacy in societal level
Sustainable Nanocellulose UV Filters for Photovoltaic Applications: Comparison of Red Onion (Allium cepa) Extract, Iron Ions, and Colloidal Lignin
No full textPublisher Copyright: © 2025 The Authors. Published by American Chemical Society.This study explores the stability of cellulose-based films as sustainable ultraviolet (UV) light filter films for optoelectronic applications. To address the gap in assessing the long-term performance of biobased UV filters in practical applications, these films were applied to dye-sensitized solar cells (DSSCs)─devices that are extremely prone to UV degradation. This research employs cellulose nanofiber (CNF) and 2,2,6,6-tetramethylpiperidine-1-oxyl radical (TEMPO)-oxidized CNF (TOCNF) based films as a basis for UV filter materials, providing the first insights into their extended reliability and functionality. The films include TOCNFs with physically and chemically physically cross-linked iron ions (III) forms (TOCNF-Fe3+ and TOCNF-ECH Fe3+), CNF film with lignin nanoparticles deposition (CNF-LNP), and CNF film dyed with red onion (Allium cepa) skin extract (CNF-ROE). UV-vis-NIR spectroscopy demonstrated that CNF-ROE blocked 99.9% of radiation below 400 nm, showcasing its superior UV-blocking capability compared to the other materials tested here. The biobased films caused a more significant loss in transmittance in the visible range than the commercial reference. Among them, CNF-ROE, which offered the highest UV protection, also demonstrated the highest light transmittance, exceeding 80% in the 650-1100 nm range. During 1000 h of light soaking testing, DSSCs covered with CNF-ROE presented minimal visual discoloration, or bleaching, of the electrolyte even compared to the cells protected by the commercial UV filter film used as a benchmark. Predictive modeling based on the accelerated aging test projected that CNF-ROE could protect DSSCs for approximately 8500 h, compared to only 1500 h with the commercial filter. To summarize, CNF-ROE stood out as a promising biobased UV filter alternative, particularly it maintained well its performance throughout prolonged exposure. The study highlights the effectiveness of biobased UV filter films for optoelectronic applications, particularly where sustainable and durable materials are paramount.Peer reviewe
Comparison, co-existence and interoperability of heterogeneous IEC 61499 control systems
No full textIndustry 4.0 has gained significant traction across manufacturing and process industries, driving the transformation of traditional automation systems into more intelligent, flexible, and decentralized control architectures. As a result, the IEC~61499 standard has attracted considerable attention, as it aligns with the core motivations of Industry 4.0 by providing a foundation for developing vendor-independent and heterogeneous distributed automation solutions. Over the past decade, several software tools and control devices compliant with this standard have emerged, each utilizing different runtime environments. This thesis explores the interoperability and co-existence of heterogeneous control systems developed using two such tools: Neptune Function Block Builder (NFBB) and EcoStruxure Automation Expert (EAE).
The study starts with a comprehensive comparison between NFBB and EAE in terms of their software architecture, runtime environments (FBSRT and UAO respectively), usability, and support for IEC~61499 compliance profiles. EAE, as a mature commercial tool, offers advanced features and broad vendor interoperability. In contrast, NFBB introduces a modern web-based interface with multi-language programming support. Despite architectural differences, both tools demonstrated basic portability through the exchange of function blocks in XML format with minimal modifications.
To validate their capabilities, both tools were applied in practical implementations on an automotive assembly demonstrator. Each tool was used to develop an independent control solution using its respective hardware controllers. Additionally, the study demonstrated system co-existence by allowing both applications to run on the same physical setup with seamless switching. Interoperability between the two systems was successfully achieved using OPC UA communication, enabling cross-platform control.
This work highlights the practical viability of integrating heterogeneous IEC~61499 systems and provides valuable insights into the strengths and limitations of emerging and established development tools
Female Entrepreneurs in the Startup Ecosystem
No full textWhile female startup leaders contribute significantly to innovation and economic growth, they remain a minority in the startup ecosystem, and limited research exists on how they conceptualize and practice leadership as well as navigate related challenges. This bachelor’s thesis systematically reviews existing studies on female startup leaders, focusing on the obstacles they face and how these shape their leadership practices and conceptualizations.
Using a systematic literature review methodology, this thesis critically analyses 44 articles from the Scopus database and citation networks, with a focus on high growth startup contexts in the Global North written between 2000 and 2025 in English. This allows for an exploration of how challenges influence female leaders in the startup context.
The review finds that women encounter systemic challenges, including funding disparities, talent attraction difficulties, and persistent gender stereotypes. These issues along with a range of barriers constrain how they are perceived, supported, and able to operate as leaders. Gender norms also place women in a “double bind”, expected to display femininity, while effective leadership is associated with masculinity. Role models emerge as important sources of motivation, stereotype challenging, and alternative leadership conceptualizations.
In terms of leadership practices, female startup leaders frequently adopt transformational, democratic, and shared styles that emphasize collaboration and relational approaches. Their motivations for founding startups often extend beyond financial metrics, prioritizing community and social impact.
The thesis concludes that systemic barriers and gendered expectations profoundly shape how women define and enact leadership in startups, leading to adaptive yet constrained approaches. Future research should empirically test these findings and further examine women’s leadership practices and conceptualizations in specific contexts, along with addressing other gaps in research. The findings of this thesis are especially relevant for female startup leaders, startup ecosystem entities and actors, as well as the government and other policy makers, guiding them on how to address problems and better support entrepreneurial women.Vaikka naisjohtajat startup-yrityksissä edistävät merkittävästi innovaatiota ja alan talouskasvua, he muodostavat vähemmistön startup-ekosysteemissä, ja tutkimusta heidän johtamiskäsityksistään, käytännöistä ja kohtaamistaan haasteistaan on rajallisesti. Tämä kandidaatintutkielma kokoaa yhteen aiempaa tutkimusta naisyrittäjistä startup-kontekstissa, keskittyen heidän kohtaamiinsa esteisiin sekä miten nämä vaikuttavat heidän muodostamiin johtamiskäsityksiin ja käytäntöihin.
Tutkielmassa analysoidaan systemaattisen kirjallisuuskatsauksen menetelmin 44 Scopus-tietokannasta ja viittausverkoista haettua artikkelia, jotka on julkaistu englanniksi vuosina 2000–2025 ja käsittelevät korkean kasvun startup-yrityksiä Länsimaissa. Näin voidaan tarkastella, miten haasteet vaikuttavat naisten johtajuuteen startup-kontekstissa.
Katsaus osoittaa, että naiset kohtaavat systeemitason haasteita, kuten rahoituksen epätasa-arvoa, vaikeuksia houkutella työntekijöitä sekä sitkeitä sukupuolistereotypioita. Nämä yhdessä muiden esteiden kanssa rajoittavat sitä, miten naisiin suhtaudutaan, tuetaan ja miten he voivat toimia johtajina. Sukupuoliroolit myös asettavat naiset ristiriitaiseen asemaan; heiltä odotetaan feminiinisyyttä, vaikka tehokas johtajuus yhdistetään maskuliinisiin piirteisiin ja toimintatapoihin. Roolimallit nousevat tärkeiksi motivaation lähteiksi, stereotypioiden haastajiksi ja vaihtoehtoisten johtamistyylien tarjoajiksi.
Johtamiskäytäntöjen osalta naiset startup-johtajina omaksuvat transformationaalisia, demokraattisia ja jaettuja johtamistyylejä, jotka korostavat yhteistyötä ja relationaalisuutta. Heidän motiivinsa yrittäjyyteen ulottuvat usein taloudellisia mittareita pidemmälle, painottaen yhteisöllistä ja yhteiskunnallista vaikuttavuutta.
Tutkielma toteaa, että systemaattiset esteet ja sukupuolittuneet odotukset muovaavat syvästi sitä, miten naiset määrittelevät ja toteuttavat johtajuutta startupeissa, johtaen mukautuviin mutta samalla rajoittuneisiin toimintatapoihin. Jatkotutkimusten tulisi empiirisesti testata näitä havaintoja sekä syventää ymmärrystä naisten johtamiskäytännöistä ja käsityksistä eri konteksteissa, samalla kun tunnistetaan ja täydennetään muita tutkimusaukkoja. Tutkielman tulokset ovat erityisen merkityksellisiä naisyrittäjille, startup-ekosysteemin toimijoille sekä hallitukselle ja muille päätöksentekijöille, tarjoten suuntaviivoja ongelmien ratkaisemiseksi ja naisyrittäjien tukemisen vahvistamiseksi
Movable Antenna-Equipped UAV for Data Collection in Backscatter Sensor Networks: A Deep Reinforcement Learning-Based Approach
No full textBackscatter communication (BC) becomes a promising energy-efficient solution for future wireless sensor networks (WSNs). Unmanned aerial vehicles (UAVs) enable flexible data collection from remote backscatter devices (BDs), yet conventional UAVs rely on omni-directional fixed-position antennas (FPAs), limiting channel gain and prolonging data collection time. To address this issue, we consider equipping a UAV with a directional movable antenna (MA) with high directivity and flexibility. The MA enhances channel gain by precisely aiming its main lobe at each BD, focusing transmission power for efficient communication. Our goal is to minimize the total data collection time by jointly optimizing the UAV’s trajectory and the MA’s orientation. We develop a deep reinforcement learning (DRL)- based strategy using the azimuth angle and distance between the UAV and each BD to simplify the agent’s observation space. To ensure stability during training, we adopt Soft Actor-Critic (SAC) algorithm that balances exploration with reward maximization for efficient and reliable learning. Simulation results demonstrate that our proposed MA-equipped UAV with SAC outperforms both FPA-equipped UAVs and other RL methods, achieving significant reductions in both data collection time and energy consumption.Peer reviewe
Rakennetut viheralueet hiilensitojina - ilmastonmuutoksen hillintää kaupunkiympäristöissä
No full text-Tämän kandidaatintutkielman tarkoituksena on selvittää, millainen rooli rakennetuilla viheralueilla on ilmastonmuutoksen hillinnässä, kuinka ne toimivat hiilensitojina ja miksi niiden hiilensidonta on merkityksellistä. Lisäksi tarkoituksena on tunnistaa keinoja, joiden avulla hiilensidonta voitaisiin paremmin huomioida rakennettujen viheralueiden suunnittelussa. Rakennettujen viheralueiden määrittelyssä on käytetty taustalla rakennetun ympäristön luontotyyppejä, ja tässä kandidaatintutkielmassa määritelmään kuuluvat kaikki ihmisen suunnittelemat kaupunkien viheralueet, lukuun ottamatta vesistö- ja kosteikkoalueita.
Ilmastonmuutos edetessään aiheuttaa vakavia haittoja ympäristössä, ja sen hillitsemiseksi tarvitaan monenlaisia keinoja, kuten hiilensidontaa. Hiilensidonnassa voitaisiin hyödyntää paremmin kaupunkiympäristöjä, kuten rakennettuja viheralueita. Tutkielmassa selviää, että rakennettujen viheralueiden rooli ilmastonmuutoksen hillinnässä globaalissa mittakaavassa voi olla melko maltillinen, mutta niiden rooli kasvaa erityisesti kaupunkimittakaavassa. Monilla kaupungeilla on esimerkiksi ilmastotavoitteita, joita voidaan tukea rakennettujen viheralueiden hiilensidonnalla, esimerkiksi osa kaupungin aiheuttamista päästöistä voidaan sitoa rakennetuilla viheralueilla. Tutkielmassa korostuu myös hiilensidontaa tehostavien toimien muutkin hyödyt, kuten biodiversiteetin ja ihmisten hyvinvoinnin tukeminen.
Tässä kandidaatintutkielmassa tarkastelu keskittyy erityisesti rakennettujen viheralueiden maaperän, kasvillisuuden ja niiden elinkaaren sekä kaupunkisuunnittelun rooliin hiilensidonnassa ja hiilivarastoissa. Tutkielmassa erityisesti maaperän rooli nousee merkittäväksi, ja kasvillisuudesta erityisesti puilla on suuri merkitys hiilensidonnassa. Tutkielmassa selvinneitä keinoja vahvistaa rakennettujen viheralueiden hiilensidontaa ja hiilivarastoja ovat esimerkiksi tiiviin, monimuotoisen ja paikallisen kasvillisuuden hyödyntäminen, olemassa olevan kasvillisuuden ja maaperän säilyttäminen, biohiilen käyttö, ja sen mahdollistaminen, että kasvillisuus voi kasvaa pitkään. Lisäksi kaupunkisuunnittelulla, esimerkiksi kaavoituksella ja numeraalisilla tavoitteilla, voidaan myös vaikuttaa rakennettujen viheralueiden hiilensidontaan ja hiilivarastoihin. Hyödynnetyistä lähteistä myös selviää, että aihe kaipaisi osittain vielä lisätutkimusta, ja tarvittaisiin parempia työkaluja kaupunkien viheralueiden ilmastovaikutusten arvioimiseen.
Tämä kandidaatintutkielma on toteutettu kirjallisuustutkimuksena, ja aineistona on käytetty muun muassa kaupunkiympäristöjen hiilensidontaan liittyviä tutkimuksia sekä Helsingin kaupungin julkaisuj
Deep learning-based metal and scatter artifact reduction in conebeam computed tomography
No full textCone-Beam Computed Tomography (CBCT) provides high-quality three-dimensional X-ray imaging and offers advantages such as reduced radiation dose, lower cost, and a smaller physical footprint compared to Multi-Detector CT (MDCT). Due to these features, CBCT is well-suited for a range of clinical applications, including dentistry, orthopedics, interventional radiology, and image-guided therapies, as well as for use in mobile clinics and remote deployments, thereby contributing to broader accessibility in healthcare. However, CBCT image quality is often compromised by inherent artifacts, including scatter and metal artifacts, which pose significant challenges to diagnostic applications.
The emergence of deep learning methods presents a promising avenue for addressing these imaging challenges, potentially offering substantial improvements. However, practical constraints, such as the need for large training datasets, and the seamless integration of deep learning models into existing artifact correction pipelines, must be addressed to ensure clinical feasibility.
The main purpose of this thesis is to develop clinically applicable deep learning techniques to mitigate metal and scatter artifacts in CBCT imaging. For instance, to overcome the challenge of data scarcity, simulated datasets are leveraged for network training. Additionally, lightweight Convolutional Neural Network (CNN) models are introduced to facilitate efficient integration into established artifact correction workflows. To ensure real-world applicability, the proposed methods are evaluated using real CBCT datasets. The research is structured around four key contributions. Publication I introduces a learning-based inpainting method of metal traces to reduce metal artifacts. Publication II employs simulated data to train a neural network for metal trace segmentation to improve the effectiveness of existing inpainting-based metal artifact reduction methods. Publication III presents a neural network for scatter estimation under clinically relevant variations in the Field of Measurement (FOM). Finally, in Publication IV, an ultrafast scatter estimation approach is proposed for deployment in mobile CBCT systems and on-device applications.
The findings demonstrate that the developed models substantially enhance the state-of-the-art in artifact correction, advancing the clinical viability of CBCT imaging through deep learning-driven lightweight solutions.navigointi mahdollistastrukturell navigationstructural navigatio
Graphlet decomposition dataset of Tallinn's road network from January 2020 OpenStreetMap data
No full textPublisher Copyright: © 2025This paper presents a comprehensive dataset of graphlet decomposition for the road network of Tallinn, Estonia, based on OpenStreetMap (OSM) data representing the road network state as of 1 January 2020. Graphlets, which are small subgraphs, serve as powerful tools for analyzing and classifying local street structures in urban networks. The dataset includes counts of all possible four-node graphlet configurations for each intersection in the road network, provided in both Comma Separated Values (CSV) and Environmental Systems Research Institute (ESRI) Shapefile formats for maximum accessibility. The methodology for extracting these graphlets using Python and the Python ORbit Counting Algorithm (PyORCA) library is explained in details. The processing pipeline includes graph construction from spatial data, node-centric graphlet counting, and conversion back to geographic format. The resulting dataset enables researchers to identify recurring patterns in urban street networks, study urban morphology, and compare structural similarities between different urban areas. The code is designed for reproducibility, allowing researchers to apply the same analysis to other cities. This dataset contributes to the growing field of quantitative urban morphology and can support studies in urban planning, transportation network analysis, sustainable development, and comparative urban studies.Peer reviewe
A risk assessment of an autonomous navigation system for a maritime autonomous surface ship
No full textThe maritime industry is undergoing a of Marine Autonomous Surface Ships (MASS). It is expected that, with the introduction of maritime autonomous technologies the safety performance of maritime operations increases. Nevertheless, the safety of the MASS must be properly planned and then demonstrated. This applies to the functionality of any Autonomous Navigation System (ANS). The nature of the risks related to an ANS might differ significantly from the traditional maritime risks. Therefore, mitigating these risks before applying an ANS to real-world operations is critical. This study analyses and assesses the risks of a currently developed ANS by integrating the System-Theoretic Process Analysis (STPA) and Bayesian Networks (BN) into a two-stage process. The study results provide a holistic view of the underlying risks in the ANS and the functionality of formulated Risk Control Options. The risk levels of loss events are computed within a model to represent the effects that each loss may have on the safety of the ship’s navigation and the company’s overall reputation. The assessed Risk Control Options provide information to system developers to elaborate an ANS that is safe by design.Peer reviewe