VTT Research System
Not a member yet
161399 research outputs found
Sort by
Multivariate analysis on simulated moisture damage emission to indoor air
No full textMoisture damage in buildings is a significant source of indoor air problems, releasing e.g. volatile organic compounds (VOCs) and microbially produced VOCs (MVOCs), which can cause unpleasant odors and health symptoms. However, interpreting MVOCs as indicators of mold is challenging due to their various sources and limitations in analytical methods. The objective of this study was to identify the most critical factors influencing VOC emissions from moisture-damaged wall structures into the indoor environment via structural air leakages. The research was conducted using the VTT Indoor Air Quality (IAQ) Simulator and analyzed with Principal Component Analysis (PCA). The IAQ simulator was used to investigate the transport of airborne impurities from mold-contaminated wall structures in realistic building conditions and the systematic manipulation of key environmental parameters. The resulting dataset was subjected to multivariate analysis to identify the most influential factors contributing to IAQ degradation in moisture-damaged structures. The key conclusions revealed that material relative humidity was the most significant single factor affecting all VOC concentrations; higher humidity consistently increased emissions. Four specific ketones (2-pentanone, 2-hexanone, 2-heptanone, and 2-octanone) were clearly identified as originating from microbial growth, with their concentrations being significantly higher in the presence of active mold growth. Pressure differentials had only a borderline effect on gypsum board emissions, while the insulation layer showed no significant impact on any of the identified VOC components. These findings underscore the critical role of relative humidity in determining indoor VOC profiles and highlight the value of multivariate methods in assessing mold-related indoor air problems.</p
Self-supervised representation learning for cloud detection using Sentinel-2 images
Full text linkThe unavoidable presence of clouds and their shadows in optical satellite imagery hinders the true spectral response of the Earth underlying surface. Accurate cloud and cloud shadow detection is therefore a crucial preprocessing step for optical satellite images and any downstream analysis. Various methods have been developed to address this critical task and can be broadly categorized in physical rule-based methods and learning based methods. In recent years, machine learning based methods, particularly deep learning frameworks, have proven to outperform physical rule-based models. However, these approaches are mostly fully supervised and require a large amount of pixel-level annotations whose obtention is costly and time consuming. In this work, we propose to deal with cloud and cloud shadow detection in optical satellite images using self-supervised representation learning, a machine learning paradigm that focuses on extracting relevant representations from unlabeled data, which can then be used as an effective starting point to fine-tune models with few labeled data in a supervised fashion. These approaches were shown to perform competitively with fully supervised methods without the requirement of large annotation datasets. Particularly, we assessed two self-supervised representation learning methods that use different philosophies about self-supervision: Momentum Contrast (MoCo), based on contrastive learning and DeepCluster, based on clustering. Using two publicly available Sentinel-2 cloud datasets, namely WHUS2-CD+ and CloudSEN12, we show that MoCo and DeepCluster, trained with only 25% of the annotated data, can perform better than physical rule-based methods such as FMask and Sen2Cor, weakly supervised methods and even several fully supervised methods. These results point out the strong applicability of self-supervised representation learning methods to the task of cloud and cloud shadow detection with self-supervised pretraining leading to fine-tuned models that outperform industry standards and achieve near state-of-the-art performances with a fraction of the data
Understanding the Gender Gap in the Acceptance of Automated Vehicles:International Mobility Study Across 17 Countries
No full textThe common assumption is that men are more likely to accept automated vehicles (AVs) than women. However, studies have produced mixed results regarding this gender gap. Additionally, there is limited understanding of how the gender gap in the intention to use AVs might vary between countries. This study aims to enhance the understanding of the gender gap in willingness to use AVs and how this gap might differ across various countries. To accomplish this, survey data from 18,631 respondents across 17 countries: Brazil, China, Finland, France, Germany, Hungary, India, Indonesia, Italy, Japan, Russia, Spain, South Africa, Sweden, Turkey, the UK, and the US, was analyzed. In this research, the gender gap in willingness to use AVs is defined as the difference in willingness to use AVs between men and women. The results indicate that gender differences in willingness to use AVs are not universal; some countries show opposing trends between men and women, while in others, the gender difference is not statistically significant. This study contributes to existing literature by examining the influence of gender and country on the willingness to use AVs. The findings have the potential to significantly impact policy development and transport planning by promoting gender inclusivity in future transport solutions, ensuring that all potential users can benefit from adopting AVs.</p
How to get the most out of fungal biotechnology?
No full textDuring the past decades, the importance of fungal biotechnology in advancing a bioeconomy and a circular economy has been emphasized in both scientific literature, project proposals, awarded grants and social media. Filamentous fungi have been proven to provide sustainable solutions for various industrial applications, ranging from bioremediation and medicine to the production of food, feed, materials, chemicals and energy. This is where we are today, but where could tomorrow’s fungal biotechnology take us? How can the seemingly infinite potential of fungal biotechnology for a circular economy become unlocked? In this editorial, we will cover some of the critical aspects that we believe are essential for the success and impact of fungal biotechnology to a future bioeconomy.</p
A Framework for Road Authorities to Assess Their Readiness to Support Connected and Automated Driving
No full textConnected and Automated Driving (CAD) will bring disruption to individuals, economies, and society. Most forms of CAD require some level of support from the infrastructure for their safe operation, in particular communications. However, additional infrastructure services to support CAD could improve safety and robustness and bring further benefits such as increased capacity. However, the infrastructure requirements of vehicle Original Equipment Manufacturers (OEMs) are not always clear, and it is therefore difficult for National Road Authorities (NRA) to prepare future levels of support for CAD, given rapidly evolving technology and uncertain projections of future CAD demand. There is a need to articulate those requirements, bringing stakeholders together to formulate a structured approach, and a roadmap that will advance safe and smart roads that support CAD. This paper presents the DiREC project (consortium partners: TRL, ARUP, TU Delft, VTT, VTI and FEHRL) funded by the CEDR Transnational Road Research Programme Call 2020 with funding provided by CEDR members of Belgium (Flanders), Denmark, Ireland, Israel, Netherlands, Norway, Sweden, Switzerland and the United Kingdom. DiREC is seeking to address the above challenge. The project has established a CAV-Readiness Framework (CRF) based on a level of service approach to understand the needs of CAD, and to define the infrastructure and services that NRAs could provide to support these needs.</p
Comparative Analysis of YOLOv8 and YOLOv9 on a Unified Traffic Sign Dataset
No full textEmploying a unique set of traffic signs, this research examines the effectiveness of two versions of the You Only Look Once (YOLO) object detection framework: YOLOv8 and YOLOv9. The aim is to provide a detailed understanding of each version’s strengths and weaknesses in traffic sign detection, with implications for enhancing real-world object detection in dynamic traffic scenarios. Preliminary findings indicate that YOLOv9 outperforms YOLOv8, demonstrating higher precision and F1-score. This highlights YOLOv9’s potential for robust traffic sign detection solutions. Despite assessment, our research represents an essential contribution to the discipline of computer vision including applications to traffic sign recognition. The study’s results offer a useful resource for practitioners and researchers selecting optimal models for similar applications, ultimately contributing to developing smarter and more efficient transportation networks.</p
How to get the most out of fungal biotechnology?
No full textDuring the past decades, the importance of fungal biotechnology in advancing a bioeconomy and a circular economy has been emphasized in both scientific literature, project proposals, awarded grants and social media. Filamentous fungi have been proven to provide sustainable solutions for various industrial applications, ranging from bioremediation and medicine to the production of food, feed, materials, chemicals and energy. This is where we are today, but where could tomorrow’s fungal biotechnology take us? How can the seemingly infinite potential of fungal biotechnology for a circular economy become unlocked? In this editorial, we will cover some of the critical aspects that we believe are essential for the success and impact of fungal biotechnology to a future bioeconomy.</p
Programming Education with LLMs and NPCs: A Dialogical Learning Framework for VS Code
No full textThe integration of Large Language Models (LLMs) holds significant promise for enhancing programming education by providing personalized, immediate feedback and fostering student engagement through adaptive learning. Existing research demonstrates that LLMs can offer meaningful learning support when designed with pedagogical considerations. However, challenges such as hallucinations in feedback and learners’ negative attitudes towards automated instructional solutions hinder broader adoption in educational contexts. Many existing solutions fail to promote critical thinking or address diverse learner needs. This paper proposes design of an interactive story-based environment for teaching programming languages, utilizing LLMs and automated assignment grading through a plug-in for Visual Studio Code (VSC). The goal is to provide contextual, pedagogically relevant tasks and feedback to students, employing Socratic questioning to encourage active participation and critical thinking. We adapt an existing VSC plug-in framework to support PHP and other common languages, designing middleware that enriches student prompts and redirects them to a custom-tuned curriculum-driven Swedish LLM. This architecture integrates meta-prompts based on pedagogical strategies into LLM interactions, employing a Socratic dialogue approach rather than providing direct answers. Anticipated outcomes include increased student engagement through storyline-based tasks and personalized feedback within the VSC environment, alongside better alignment of LLM interactions with pedagogical objectives. By presenting the underlying architecture of the prototype, we contribute to the use of generative AI in software engineering education. Our work highlights the potential of AI-powered tools in education to improve learning while addressing ethical considerations and ensuring need for thoughtful implementation to avoid amplifying biases or diminishing the role of teachers. Further studies are recommended to evaluate the impact of LLM interactions on student learning outcomes and to explore adaptability in real-life educational simulations
Direct-to-Device Connectivity for Aviation:Opportunities for Integrated CNS Services
No full textSatellites are a key component of aeronautical telecommunication networks for supporting communication, navigation, and surveillance (CNS) services. Satellite communication, also known as SATCOM, acts as a bridge between aircraft and terrestrial infrastructure and establishes air-to-ground data (A2G) datalinks to connect aircraft with the air navigation service providers. For example, SwiftBroadband-Safety (SB-S) has been emerged to provide a global, secure, broadband IP connection for both operations and safety communications to aircraft, which can support CPDLC and ADS-C services with the same safety services, helping airlines to be ready for future air traffic management evolutions. Iridium, partnership with Aireon, supports satellite-based automatic dependent surveillance – broadcast (ADS-B) in oceanic areas. Global navigation satellite system (GNSS) is the underlying technology that enables safe navigation and provides precise location data for ADS-B [1].As an important milestone, Airbus and OQ Technology through their fruitful collaboration have demonstrated the feasibility of connecting an unmanned aircraft, carrying a 5G user equipment, to a low-Earth orbit (LEO) satellite running a full stack of 5G base station. Recently, many major airlines started to offer Starlink connectivity to passengers enabling Internet services including browsing and 4K video streaming. Despite these advancements, small unmanned aircraft may not be able to benefit from traditional satellite systems, whose terminals are often large and energy-hungry.Direct-to-Device (D2D) connectivity is an emerging concept in new space era satellite communications [3]. D2D connects compact consumer devices e.g., smartphones, wearables, and machine type device directly to Earth orbiting satellites without relying on terminal or mediator gateways. Due to small device form factors and high energy efficiency, D2D appears as a promising solution for meeting the CNS service requirements of unmanned aircraft system (UAS). Particularly, D2D devices due to small form factor and low energy consumption can be mounted on the unmanned aircraft and could provide satellite-based A2G links to support CNS services. D2D links can enable beyond-line-of-sight coverage over remote and oceanic regions, augments GNSS accuracy through correction data and 5G NTN support, and facilitates transmission of ADS-B and ADS-C, enabling situational awareness,and resilient CNS operations for UAS