Occupational safety and health is traditionally a challenging area in the labor-intensive construction industry as accidents at work and non-ergonomic work conditions lead to absences and premature retirement of construction workers. Recently, the rise of the Internet of Things (IoT) and its accompanying technologies (e.g. wearable technologies) has enhanced interest in the occupational safety and health of construction work. The level of technology acceptance among construction workers is a crucial element in the adoption of these technologies. The main objectives of this study are to enhance understanding about construction workers’ attitudes towards IoT-based data-intensive work safety and wellbeing solutions and to identify factors that can promote technology adoption. Data for the study was collected through an online survey of 4385 construction workers. Based on the survey data it seems that construction workers would accept the sharing and utilizing data collected from them in the worksite environment if it could help identify employee personal health risks or promote personal and colleagues' occupational safety. Respondents were most concerned about privacy and security regarding wearables in the workplace. It can be concluded that user acceptance and trust building are key components in the adoption of IoT-based occupational safety and health solutions. Future studies should investigate methods for actively involving construction workers in the design and development process of IoT-based work safety solutions and examine technological solutions that promote trust building among construction workers.

Sustainable work aims at improving working conditions to allow workers to effectively extend their working life. In this context, occupational safety and well-being are major concerns, especially in labor-intensive fields, such as construction-related work. Internet of Things and wearable sensors provide for unobtrusive technology that could enhance safety using human activity recognition techniques, and has the potential of improving work conditions and health. However, the research community lacks commonly used standard datasets that provide for realistic and variating activities from multiple users. In this article, our contributions are threefold. First, we present VTT-ConIoT, a new publicly available dataset for the evaluation of HAR from inertial sensors in professional construction settings. The dataset, which contains data from 13 users and 16 different activities, is collected from three different wearable sensor locations.Second, we provide a benchmark baseline for human activity recognition that shows a classification accuracy of up to 89% for a six class setup and up to 78% for a sixteen class more granular one. Finally, we show an analysis of the representativity and usefulness of the dataset by comparing it with data collected in a pilot study made in a real construction environment with real workers.

In this paper we propose an unsupervised speaker change detection (SCD) system developed for mobile device applications. We use Bayesian information criterion (BIC) to find initial speaker changes, which are then verified or discarded in the second phase by utilizing modified BIC and silence detector information. Silence information usage after initial BIC in decision making is useful to separate real changes from noise peaks. Enhanced peak detector adjusts BIC penalty parameter automatically, which improve the robustness and feasibility. Improved BIC based false alarm compensation (FAC) merges effectively consecutive segments belonging to same speaker. Our experiments have shown the robustness of the algorithm and it produces very satisfactory results for difficult mobile phone recorded speech data.

This dataset presents the first open-access collection of near-infrared hyperspectral imaging (NIR-HSI) data for the optical identification of textiles, with a focus on supporting research in sensor-based textile sorting and recycling. The dataset comprises hyperspectral images, RGB photographs, and detailed metadata, including fibre composition and colour, for 71 post-industrial textile samples, collected in Finland. Over 11 million spectra are included in the hyperspectral images, with more than 6 million annotated, providing a robust foundation for machine learning and data analysis. In addition, we provide a single representative NIR spectra and RGB value for each sample in order to accommodate classic spectroscopic analysis.

Background: Pervasive technologies are used to investigate various phenomena outside the laboratory setting, providing valuable insights into real-world human behavior and interaction with the environment. However, conducting longitudinal field trials in natural settings remains challenging due to factors such as low recruitment success and high dropout rates due to participation burden or data quality issues with wireless sensing in changing environments. Objective: This study gathers insights and lessons from 3 real-world longitudinal field studies assessing human behavior and derives factors that impacted their research success. We aim to categorize challenges, observe how they were managed, and offer recommendations for designing and conducting studies involving human participants and pervasive technology in natural settings. Methods: We developed a qualitative coding framework to categorize and address the unique challenges encountered in real-life studies related to influential factor identification, stakeholder management, data harvesting and management, and analysis and interpretation. We applied inductive reasoning to identify issues and related mitigation actions in 3 separate field studies carried out between 2018 and 2022. These 3 field studies relied on gathering annotated sensor data. The topics involved stress and environmental assessment in an office and a school, collecting self-reports and wrist device and environmental sensor data from 27 participants for 3.5 to 7 months; work activity recognition at a construction site, collecting observations and wearable sensor data from 15 participants for 3 months; and stress recognition in location-independent knowledge work, collecting self-reports and computer use data from 57 participants for 2 to 5 months. Our key extension for the coding framework used a stakeholder identification method to identify the type and role of the involved stakeholder groups, evaluating the nature and degree of their involvement and influence on the field trial success. Results: Our analysis identifies 17 key lessons related to planning, implementing, and managing a longitudinal, sensor-based field study on human behavior. The findings highlight the importance of recognizing different stakeholder groups, including those not directly involved but whose areas of responsibility are impacted by the study and therefore have the power to influence it. In general, customizing communication strategies to engage stakeholders on their terms and addressing their concerns and expectations is essential, while planning for dropouts, offering incentives for participants, conducting field tests to identify problems, and using tools for quality assurance are relevant for successful outcomes. Conclusions: Our findings suggest that field trial implementation should include additional effort to clarify the expectations of stakeholders and to communicate with them throughout the process. Our framework provides a structured approach that can be adopted by other researchers in the field, facilitating robust and comparable studies across different contexts. Constantly managing the possible challenges will lead to better success in longitudinal field trials and developing future technology-based solutions.