Chemical accidents in industrial parks and chemical plants often involve cascading events and multiple interacting hazards that evolve rapidly, challenging traditional evacuation planning for complex technological systems. These hazards include fires, explosions, and toxic releases. To understand how evacuation can be planned under these multi-hazard dynamics, this paper follows PRISMA 2020 guidelines, systematically screens database records using predefined inclusion or exclusion criteria, and synthesizes 95 included studies. This review synthesizes how accident consequence/risk outputs (e.g., tenability, dose, and probability metrics) are translated into evacuation-model inputs, consolidating a five‑level integration staircase (Levels 0–4) describing increasing coupling between risk assessment and evacuation planning, from static zoning (L0) and fixed, risk‑weighted routes (L1) to time‑dependent optimization (L2), physics–crowd co‑simulation (L3), and sensor‑assisted re‑planning under uncertainty (L4). This paper also synthesizes four priorities for future work: rapid and reliable hazard forecasting; real‑time, sensor‑informed re‑planning; behaviorally realistic crowd and congestion modeling; and decision frameworks that explicitly quantify uncertainty to support auditable decision support systems. Together, these elements outline a staged, auditable pathway from offline pre‑plans to online, explainable, risk‑aware operations for facilities and nearby communities.

Poor indoor air quality (IAQ) is directly linked to respiratory and cardiovascular diseases, allergies, and other health conditions. Mold significantly contributes to poor IAQ, which is closely related to various health issues, including respiratory and cardiovascular conditions. Raising public awareness about IAQ is essential for empowering individuals to adopt preventive measures. This study aims to develop a serious game designed to educate the general public on preventing mold growth in residential buildings in New Zealand. The game, developed using Articulate Storyline 360, engages users in identifying mold-prone areas in kitchens, bedrooms, and bathrooms, understanding mold formation factors, optimal temperature and humidity ranges, and applying effective mitigation strategies. A semi-structured interview involving 12 expert participants was conducted to validate the game’s content. Participants highlighted key mold-prone areas, including furniture near exterior walls, ceiling corners, and curtains, and emphasized the importance of managing moisture and cold surfaces to prevent mold growth. They recommended practical strategies such as wiping condensation, using extraction fans, and maintaining indoor humidity levels between 40– 60% to mitigate mold risks effectively.

Effective fire safety training is critical to mitigating property losses and human injuries and deaths. Augmented Reality (AR) is an emerging solution for safety training, offering a dynamic and interactive learning environment. Despite the emergence of various AR training prototypes demonstrating potential advantages, there is limited research on how teaching theories impact the efficacy of augmented reality. This study aims to integrate and test two teaching theories, directive instruction and productive failure, into AR-based fire safety training. Two novel AR prototypes, directive instruction training and productive failure training, were designed and prototyped in this work. These two prototypes were tested in a controlled experiment involving 68 participants. A comparison was carried out, focusing on knowledge acquisition, knowledge retention, intrinsic motivation, and self-efficacy. The results indicate that productive failure training has better performance in enhancing knowledge acquisition and knowledge retention. However, both prototypes perform equally in the overall learning experience. As such, these findings offer valuable insights into future augmented reality development for safety training.

The art of transferring knowledge from one source to another is how humankind has built most of the vast wisdom we enjoy today. Knowledge sharing occurs at much higher rates today due to new digital ways of capturing and sharing knowledge. Recent literature suggests that extended reality-based training can benefit organizations. This paper explores the capabilities of Augmented Reality (AR), a subset of extended reality, in transferring technical knowledge within an organization. It is not well understood how the benefits of AR based training change depending on the complexity of the task. We attempt to address this gap by developing HoloLens2 based authored AR guides for four maintenance tasks of differing complexity and testing with twenty-three participants of differing levels of experience. Each participant, recruited from the Royal New Zealand Navy, performed two simple and two complex tasks. The participants used AR guides for one simple and one complex task. They completed the remaining two tasks with traditional paper instructions. The training effectiveness is assessed by time of completion of tasks, count of errors, and participants’ cognitive load. The participants using AR made fewer errors and reported a lower cognitive load for the training. However, the time spent on the AR training was longer for three of the four tasks.

Immersive technology, including virtual reality (VR), augmented reality (AR), and mixed reality (MR), is now widely applied across various fields. However, performing tasks and solving problems in immersive environments requires a certain level of digital literacy – the ability to access, evaluate, and effectively use digital technology and information. Uncovering the relationships between digital literacy and the outcomes of immersive technology usage can contribute to designing more inclusive and responsive immersive systems, while also supporting efforts to reduce digital inequalities across different contexts. As such, this study conducted a systematic literature review to investigate how digital literacy affects the use of immersive technology, including the impacts, their forms, and underlying mechanisms. The results indicate that multiple interpretations of digital literacy exist within the literature and are commonly assessed by subjective measurements. Most studies focus on education and tourism, with VR and AR being the most studied types of immersive technology. The impacts of different levels of digital literacy on the use of immersive technology can be categorised into the sense of presence, experience, affection, behaviour and performance, with the effects being positive, negative, or neutral. The impacts on behaviour and affection all show a robust positive trend, whereas the effects on sense of presence and experience are mixed and sometimes contradictory across the reviewed studies. Four potential underlying factors, namely cognitive load, affection, expectation, and operational skills, are identified, offering possible explanations for how digital literacy influences the use of immersive technology. Findings can inform future research aiming at improving the design, development, and implementation of immersive technology, thereby helping to overcome the barriers posed by digital literacy.

This study presents a synthetic dataset designed to support the validation of joint-level trajectory extraction for pedestrian dynamics measurements. Existing pedestrian datasets commonly provide pedestrian-level trajectories, but rarely include frame-wise 3D body-joint coordinates that can be used to evaluate joint-level trajectory extraction and microscopic step measurements. To address this gap, this study developed a dataset containing 147 rendered video sequences, comprising 18,669 video frames in total, of a single animated pedestrian walking along a straight path at three walking speeds. The same walking motions were rendered from 49 virtual cameras placed at different distances, heights, and viewing angles, enabling investigation of camera configuration effects. For each video, the dataset provides frame-wise ground-truth 3D body-joint coordinates in both the world and camera coordinate systems, together with camera intrinsic and extrinsic parameters. By combining rendered videos, known camera parameters, and frame-wise body-joint coordinates, this dataset provides a controlled benchmark for developing and validating vision-based methods for extracting 3D joint-level pedestrian trajectories and deriving step measurements from video. The controlled variation in camera configurations can provide guidance for camera placement in real-world pedestrian experiments.

Accurate and resilient monitoring of construction projects remains challenging due to fragmented reporting, data uncertainty and delayed system integration. This paper evaluates RealCONs, a QR-enabled real-time monitoring framework that integrates BIM, mobile scanning, cloud-based SQL storage, and Power BI analytics to support live project control. A 90-day comparative case analysis of two concurrent Electrical and Instrumentation projects benchmarked RealCONs against a conventional tracking system. Performance was assessed using Earned Value and Earned Schedule metrics, supported by Chi-square and two-proportion tests, confidence intervals, normality testing, regression forecasting, and non-parametric Wilcoxon and Mann-Whitney analyses. Data continuity strongly favoured RealCONs, with five missing earned-value days compared with 35 in the comparator project (chi(2) = 28.93, p < .001). Across 51 paired days, RealCONs achieved superior CPI (1.02 vs 0.90) and SPI (1.01 vs 0.89). During a delay event (Days 33-37), RealCONs maintained measurable progress and statistically significant SPI predictability, while the comparator recorded zero earned value. Overall, RealCONs enabled earlier delay detection, improved forecast reliability and scalable, real-time decision support aligned with Industry 4.0 objectives.

The growing frequency of wildfires poses serious threats to communities in wildland-urban interface regions. Understanding evacuation behavior is critical for effective emergency planning. This study analyzes evacuation during the 2025 Palisades and Eaton Fires using high-resolution Facebook data. We propose a framework to derive wildfire evacuation-related metrics, including population displacement rate, departure timing, delay, origin-destination flows, travel distance, and destination types. A continuous Damage-Evacuation Disparity Index (DEDI) integrates structural damage with population displacement rate to identify vulnerability hotspots. Results show displacement rate ranging from 0% to 95%, with approximately 40% of evacuations occurring within the first day. DEDI was positively correlated with the Social Vulnerability Index (correlation = 0.563). Residential areas dominated evacuation destinations (73.8-75.4%), while hotel destinations were associated with longer travel distances. Facebook-derived metrics were crosschecked with an independent GPS dataset and showed similar patterns. These findings demonstrate the Facebook data's potential for data-driven wildfire evacuation planning.

Digital Twins (DTs) hold transformative potential to enhance resilience and efficiency in water systems, yet their implementation remains fragmented across sectors and geographies. Despite growing technological momentum, the absence of consolidated evidence on practical maturity and sustainability impacts continues to limit broader adoption. This systematic review evaluates DT applications in sustainable water management, using the PRISMA framework to address these critical gaps. Quantitative analysis reveals a predominant focus on urban water infrastructure (62.5

Understanding exit choice behavior during evacuations is critical for enhancing safety and efficiency in real-life emergencies. Individual, social and environmental factors can significantly influence group decision-making during evacuations. While a substantial amount of research has focused on individual exit choice behaviors, very few studies have investigated how group decision-making is affected by various factors during emergencies, especially when group splitting occurs. This study investigates the impacts of individual, social, and environmental factors on both individual and group exit choice decision-making, using a multi-user virtual reality (VR) setup that simulates fire evacuation scenarios. We conducted experiments with 127 participants, organized in groups ranging from 1 to 4 group members. We observed a clear tendency for 2-, 3-, and 4-person groups to split into smaller units during evacuations, with 2-person subgroups being the most common and larger groups showing a higher likelihood of splitting. Then, a discrete choice model was used to analyze the effects of various social, individual and environmental factors on the participants’ exit choices. The results showed that participants were not only influenced by the factors significant in individual decision-making, such as distance to exits, familiarity with exits, and smoke, but also by the choices of other group members, demonstrating a strong social influence. Additionally, the impact of smoke was more pronounced in subgroup settings. This research highlights the significant differences between decision-making at the individual and subgroup levels during fire evacuations, providing insights for developing more sophisticated evacuation simulation models and building safety management protocols.

An understanding of how evacuees travel during bushfires remains limited, with existing research offering little consensus on the key influential factors. Additionally, bushfire studies often overlook key psychological and social (i.e., unobservable or latent) factors that may serve as mediators in evacuation travel decisions. Drawing on seminal theories commonly used in non-emergency transportation research, this study proposes a novel theoretical framework explaining destination and mode choices in bushfire evacuation, highlighting the role of observable and latent factors in the choices made. Qualitative data from 52 semi-structured interviews with fire survivors from three Australian locations, conducted between May and August 2024, have been analysed. Findings reveal new insights for bushfire evacuation literature, including the impact of attitudes (e.g., affordability, emotional support), familiarity, social connections (e.g., strength of ties), and perceived behavioural control on destination choices; as well as links between mode choices and perceived social responsibility, social connections, affect, and attitudinal factors of comfort, safety, and flexibility in travel. By addressing the omission of latent factors in current bushfire studies, this study enables the development of more comprehensive empirical models predicting bushfire evacuation travel behaviour, and in turn, improved evacuation simulation tools, community-wide evacuation plans, and real-time evacuation decision-making during fires.

To enhance the safety of building fire evacuations by accounting for the dynamic and cumulative impacts of hazardous fire conditions on evacuees, we developed SafeStep, a novel reinforcement learning-based path planning model that integrates occupant tenability into evacuation decision-making. Specifically, the model employs the Fractional Effective Dose (FED) of toxic gases to quantify dynamic fire impacts, and a FED-derived reward function is formulated to guide the RL agent toward safer and more efficient paths. A Deep Q-Network (DQN) is adopted to optimize evacuation path planning within complex building geometries. SafeStep is benchmarked against traditional path planning algorithms, Dijkstra’s algorithm (DA), through two test cases. In one case, the results show that SafeStep provides safer evacuation paths, achieving approximately a 62% reduction in FED exposure compared to DA. In the other, it successfully identifies viable evacuation routes in scenarios where the DA-based model fails. To further assess its applicability, a case study in a complex building geometry shows that SafeStep can consistently generate evacuation paths with lower FED across arbitrary starting positions. These findings indicate that SafeStep effectively addresses key limitations of traditional path planning algorithms, which often fail to account for evolving fire dynamics and cumulative fire effects. As such, the proposed model has strong potential to support smart building technologies, such as dynamic directional exit signs, to enhance evacuation safely and efficiently in real-world fire emergencies.

With the rapid development of computer graphics and hardware devices, immersive virtual reality (VR) systems designed for certain tasks such as gaming, education and training have been applied widely in our daily lives. However, further research is needed to determine whether a VR system with specific tasks (task-oriented VR) has sufficient ecological validity to achieve their intended goals. Immersion is a critical index used to evaluate the perceived verisimilitude of VR systems and user engagement, both of which are key factors of ecological validity. However, the exact operational definition of immersion and its prerequisites remain unclear and debatable. In this study, we propose and validate a new immersion scale for task-oriented VR experiences, named Virtual Reality Immersion Scale (VRIS). The initial items for the scale were developed based on a thorough review of relevant literature and were refined through a focus group, expert consultations, and a pre-survey of VR users. The resultant pre-test version of the scale was assessed and further refined using data collected from 308 participants across five task-oriented VR scenarios and from 103 participants in a fire evacuation experiment. The formal scale was then finalized through item analysis, internal consistency reliability measures, concurrent validity analysis, exploratory factor analysis, and confirmatory factor analysis. To further investigate the relationship between three factors of the VRIS, user characteristics, and VR experience types, we employed K-means clustering, multivariate logistic regression, and the Kruskal-Wallis test. The results indicated that the VRIS could effectively assess immersion in task-oriented VR experiences, providing a valuable tool for verifying the ecological validity in VR-based psychological research, and testing commercial VR experiences under development.

Knowledge of household evacuation behavior in wildfires is critical to creating effective emergency plans. Understanding evacuation route choices is especially important for both evacuation management strategies and traffic simulations during emergencies. Existing studies of evacuation route choice rely solely on survey or interview data, which have inherent limitations in spatiotemporal resolution and memory bias. This study proposes a new methodology to analyze evacuation route choices and complement existing data sources by leveraging a GPS dataset. The proposed methodology includes a framework for systematically handling sparse GPS data and testing common assumptions used by existing evacuation simulations, i.e. whether evacuees take the shortest path, make a single trip to their destination, or are at home at the start of the evacuation. We applied this new method to a sample of 155 evacuees from the 2021 Marshall Fire in Boulder County, CO, which resulted in the evacuation of over 30,000 people. We found that the majority of evacuees approximated the shortest path (67.1%), made multiple trips while evacuating to their final destination (64.7%), and were home at the start of the evacuation (60%). The findings of this study can be used to inform better plans for future emergencies and enhance traffic simulations of evacuation behavior.

Introduction: The rapid change in climate and advancement in new materials and technologies are reshaping risk landscapes with more frequent and severe hazards, emergencies, and disasters. Spatial proximity to those hazards critically shapes risk perception, influencing safety decisions and behaviors. Though a few studies have investigated how people perceive and respond to risks based on different spatial exposure to hazards, the role of spatial distance in shaping risk perception remains conceptually fragmented. This systematic review aims to: (a) evaluate methodological approaches to distance and risk perception assessment; (b) examine human factors mediating proximity effects; and (c) identify patterns in distance-risk perception relationships. Method: A PRISMA-guided analysis of 54 studies from Scopus and Web of Science was conducted to examine the literature. Results: This review identified three distance-risk perception patterns: (a) increased risk perception with proximity (45 studies), largely attributed to sensory salience; (b) reduced perception near hazards (6 studies), linked to habituation; and (c) non-linear patterns (3 studies), influenced by familiarity and motivational trade-offs. The reviewed studies applied different methods to categorize and measure the distance to hazards, including Euclidean distance, zone-based classification, and real-time sensing. Risk perception was evaluated through diverse methodologies such as surveys, validated Likert-type scales, behavioral observations, and technology-driven tools like virtual reality simulations and physiological monitoring. This review also finds that human factors, such as age, gender, education, income, and prior experience, moderate proximity effects, with older adults and women exhibiting stronger sensitivity. Practical applications: This study contributes a unified overview of methodological variation and perceptual outcomes, offering new insight for risk communication, policy design, and hazard management.

Pedestrian and crowd dynamics involves multiple disciplines, including computer science, engineering, mathematics, physics, bio-mechanics, psychology, social science and more. For effective collaboration between disciplines, researchers need a common understanding of key concepts. To address this challenge, A Glossary for Human and Crowd Dynamics was published six years ago, providing researchers with a valuable reference for cross-disciplinary communication. We now present the second version, which includes 53 new concepts and 12 revisions from the first glossary, collaboratively developed by 65 contributors from various disciplines and regions around the world through a multi-stage process. This process involved identifying new concepts not covered in the first glossary and suggesting revisions to existing entries, voting on proposed additions and modifications, writing definitions for the selected concepts, and collaboratively revising and editing the entries. By introducing new terms and refining existing definitions, this glossary aims to facilitate clearer communication, improve conceptual consistency, and support collaboration among researchers working within the field of human and crowd dynamics from diverse perspectives.

Digital Twin (DT) technology has emerged as a promising solution for enhancing the management of the three water infrastructures (potable water, wastewater and stormwater). Its application ranges from asset monitoring and predictive maintenance to health management and decision support, promising a future of efficient and sustainable water management. In New Zealand, where ageing infrastructure and climate change present significant challenges, implementing DT on the three waters sector offers substantial benefits. This study comprehensively reviews DT applications in New Zealand’s three waters infrastructure by examining academic literature, industry insights, and government initiatives. The research analyses the architecture layers of DT system using Six-layered Architecture for Digital Twin (SLADT) framework and explores stakeholders’ perspectives on DT adoption in the sector. The findings reveal DT systems follow a hierarchical framework, with layers tailored to stakeholders’ needs. Since 2018, the New Zealand government has initiated several programs, including the National Digital Twin program, to promote DT technology. However, the progress has been hindered by shifting government priorities and a lack of datasharing among stakeholders. While some councils have implemented DT at various levels, the full-scale adoption of DT across the three waters system remains incomplete. The study identifies stakeholders’ reluctance to collaborate as a key challenge, which could be addressed through legislative reforms, policy changes, and increased government funding.

Construction projects increasingly rely on processing vast amounts of data from multiple sources, including consultants (BIM), cloud-based project management platforms (e.g., Aconex), planning departments, construction sites, main contractors, and subcontractors. However, inefficiencies in data acquisition and reliance on manual data entry hinder real-time project analysis, delay notifications, and decision-making. This study introduces the Real-Time Data-Driven Construction Project Analysis Framework (RealCONs) to address these challenges by streamlining data flow and enhancing project performance. A comparative analysis used eight case studies four employing the existing approach and four utilising RealCONs—to assess improvements in data integration, early delay identification, and decision-making efficiency. The results, validated through Earned Value Management (EVM) and Earned Schedule Management (ESM) metrics, demonstrate that RealCONs significantly enhance project forecasting accuracy, schedule adherence, and cost management. Additionally, statistical analyses, including the Shapiro-Wilk test and the Wilcoxon Signed-Rank analysis, confirm that RealCONs outperform the existing approach by reducing data collection and decision-making delays, enabling project managers to implement proactive mitigation strategies. These findings highlight RealCONs’ potential to improve project efficiency, reduce costs, and optimise real-time construction management.

Assessing evacuation time is a fundamental task in fire engineering. One of the key decisions made in evacuation dynamics is exit choice. In this work, we propose a new immersive virtual reality (VR) experiment to assess the effects of social influence and fire wardens’ instructions on the exit chosen. We also investigate if and how the perceived level of authority of the fire wardens (i.e., metro staff members or firefighters) can affect these decisions. The proposed immersive VR experiment includes 12 different scenarios during a fire evacuation in an underground metro station. A sample of 131 participants took part in the experiment, making 1048 choices. We estimate a discrete choice model to quantify if and how these factors affect the participants’ decisions. The results show that both instructions by fire wardens and social influence significantly affect exit choice and that the impact of fire wardens can change depending on their perceived level of authority.