ABSTRACT In rapidly transforming cities, the quality of pedestrian environments has become central to sustainable and equitable mobility. Wayfinding signage, as a highly adaptable and comparatively low‐cost intervention, plays a critical yet frequently undervalued role in shaping how people interpret, traverse, and ultimately experience urban space. This review aims, through a Human Factors and Ergonomics perspective, to clarify the mechanisms through which wayfinding signage influences pedestrian walking perception. A PRISMA‐based systematic review identified 32 empirical studies published between 2010 and 2024. Using a structured coding strategy, signage characteristics were synthesized into recurrent human‐centered mechanisms and the perceptual outcomes they influence. The evidence portrays a field that is methodologically sophisticated but contextually narrow: research clusters around hospitals, transit hubs, and other controlled interiors, with limited attention to open streets, culturally diverse users, or groups with heightened vulnerability. Across heterogeneous methods, four HFE pathways consistently emerge—cognitive load, situation awareness, decision‐making, and emotional engagement—jointly shaping four stable dimensions of walking perception: continuity, connectivity, accessibility, and environmental attractiveness. These cross‐study regularities point to three actionable directions for design practice: strengthening perceptual clarity to reduce cognitive burden, integrating cultural and experiential meaning to enrich environmental interpretation, and building inclusive information structures that expand accessibility for diverse pedestrian groups. By consolidating dispersed findings into a coherent HFE‐informed framework, this review reframes wayfinding signage as an active interface embedded within the pedestrian experience, offering a conceptual foundation for future causal modeling and evidence‐driven urban design.
ABSTRACT The transition toward the Single Pilot Operations (SPO) in civil aviation necessitates advanced tools to manage pilot cognitive overload during task‐intensive flight phases. This study proposes a novel dynamic cognitive workload assessment framework to address multitasking demands during the approach phase. Grounded in Time‐stochastic Petri nets (TSPN) and Multiple Resource Theory (MRT), the research innovatively integrates a weighted VACP quantification method with the QN‐ACTR cognitive architecture, establishing a “Temporal‐Resource‐Cognitive” triple‐coupled model. Specifically, the framework utilizes TSPN with inhibitor arcs and stochastic time delays to formally characterize task concurrency and priority scheduling. By incorporating the decay factor from QN‐ACTR, the model captures the “reactivation costs” associated with postponed tasks resulting from working memory degradation during cognitive queuing. To validate the model, a flight simulation experiment was conducted with 29 participants across four workload intensity levels. Experimental results demonstrate that the model‐predicted values are highly correlated with NASA‐TLX scores and sensitive eye‐tracking parameters, significantly outperforming traditional static assessment methods. This dynamic framework provides a reliable quantitative tool for HMI optimization and intelligent flight system development, offering substantial practical value for enhancing aviation safety in future SPO model cockpits.
Industry 4.0 is forcing thxike Ergonomics and Human Factors (E&HF) discipline to evolve. E&HF researchers and practitioners must now grapple with features of complexity, such as dynamism, systemic emergence, and scale-features the conventional Human Reliability Assessment (HRA) toolkit is not well adapted to. This paper shows how the capability of HRA can be significantly enhanced to directly confront the long-standing challenge of identifying and measuring emergence in work systems. The NEXUS framework is introduced as a proof of concept that integrates three established approaches, agent-based modeling, DynEAST, and HRA, and applies them to a live rail maintenance case study. By embedding human error probabilities (HEPs) into a stochastic ensemble of DynEAST network simulations, NEXUS captures the divergence between deterministic baseline performance and the distribution of outcomes arising from human performance variability, providing a formal, measurable index of emergence. Comparing ensembles parameterized with observed and intervention HEPs then yields a principled basis for evaluating how effectively targeted interventions attenuate emergent variance at the system level. The findings demonstrate that NEXUS can capture meaningful emergent aspects of work system performance and test the relative impact of proposed interventions, expanding the analytical boundaries of HRA toward methods compatible with the systems thinking paradigm that now predominates in E&HF.
This study investigates whether the uncanny valley effect exists when voice assistants (VAs) have both human-like appearances and perceived minds and examines the roles played by humanness, eeriness, attractiveness, and emotional responses in this context. We draw on mind perception theory to decompose mind perception into perceived agency (autonomous capabilities) and perceived experience (emotional/cognitive capacities), which is paired with a 2 (perceived agency) & times; 2 (perceived experience) & times; 2 (appearance: human-like vs. machine-like) experimental design involving 34 participants. The results reveal that perceived experience significantly influences perceptions of humanness, eeriness, and attractiveness, whereas perceived agency has no main effect. Human-like appearances enhance humanness and attractiveness but also amplify eeriness, aligning with classic predictions associated with the uncanny valley effect. Mediation analysis reveals that humanness positively predicts both eeriness and attractiveness, and attractiveness drives positive emotional responses (pleasure and arousal), whereas eeriness has no direct effect on emotions. This research extends the uncanny valley framework to encompass mind perception attributes. The findings highlight the need for balanced anthropomorphism in VA design, suggesting caution in endowing VAs with human-like emotional experiences to mitigate eeriness while leveraging attractiveness to enhance user engagement. These insights contribute to human-computer interaction theory and offer practical guidelines for the development of user-centered intelligent systems.
This paper introduces a pattern-based analytical framework to understand how joint cognitive systems adapt within complex socio-technical environments. Building on Cognitive Systems Engineering (CSE) and the Work-as-X (WAx) framework, it formalizes work system patterns via the Structured Exploration of Complex Adaptations (SECA) method. The approach integrates data analytics with narrative inquiry, transforming qualitative accounts into structured, reusable knowledge. The resulting pattern library is expected to support proactive safety management and organizational learning at large, enhancing adaptive capacity and resilience across dynamic operational contexts.
Prolonged sedentary behavior in office settings has become a critical public health concern, contributing to various chronic diseases. With the advancement of intelligent office systems, real-time monitoring of posture and physiological responses offers new opportunities to improve workplace health; however, the relationship between seat pressure distribution and physical fatigue during extended sitting remains insufficiently explored. Here, we systematically examine seat pressure distribution characteristics and their association with cumulative physical fatigue during prolonged office sitting to support the development of intelligent office furniture with monitoring and intervention capabilities. A controlled experiment was conducted in which 18 participants (aged 23-27) completed a simulated 3-h office sitting session. Subjective fatigue was assessed every 15 min using the Borg scale, and cumulative fatigue was quantified using the area under the curve (AUC). Seat pressure distribution was continuously recorded via a pressure sensor mat, and both pressure magnitude-related indicators (e.g., mean and peak pressure) and center-of-pressure (COP)-based spatiotemporal features were extracted. Temporal trends were described, and linear mixed-effects models were applied to examine associations between objective pressure indicators and subjective fatigue while accounting for repeated measurements, sitting duration, and individual-level covariates. Subjective fatigue increased progressively over time, with distinct differences in cumulative fatigue across body regions. Several COP-based indicators showed temporal patterns broadly consistent with fatigue progression; however, sitting duration was the dominant predictor of subjective fatigue. After controlling for time and individual differences, most pressure magnitude-related and COP-based indicators showed limited independent associations with fatigue ratings. These findings indicate that although seat pressure features reflect time-dependent changes accompanying prolonged sitting, their ability to independently represent subjective physical fatigue is constrained. The results highlight the importance of time-aware analytical frameworks when interpreting pressure-based metrics and provide insights for developing interpretable fatigue monitoring approaches in intelligent office environments.
Designing tomorrow's maintenance systems with a human-centered approach is crucial to ensure optimal safety and performance. A key prerequisite for achieving this goal is to anticipate operators' cognitive behavior early in the design cycle. This study aims to determine whether the combination of mental workload measurement tools, including subjective, behavioral, and physiological measures, can detect comparable levels of cognitive effort during helicopter maintenance tasks in both real-world and virtual reality conditions. We analyzed data from 10 participants who performed four maintenance tasks of varying complexity on a helicopter, including component removal and installation. These efforts were measured using subjective scales (NASA-TLX), performance indicators (completion time), and cardiovascular data (heart rate, heart rate variability). Our observations revealed similar completion times and higher NASA-TLX scores for complex tasks, regardless of the real and virtual environment. Regarding cardiovascular data, the time-domain heart rate variability indicators showed consistent trends across both real and virtual environments. in both real and virtual settings. This research marks a significant step forward in the multidimensional, anticipatory measurement of mental workload in maintenance within a realistic industrial context.
Research has shown that flow experiences among groups, such as sports teams and music groups, can positively affect group processes and outcomes. However, there has been limited research on the effects of team flow in the workplace, and existing studies have mainly employed a 37-item questionnaire, namely the Team Flow Monitor (TFM), as a basis for measuring team climates and coaching the teams to achieve higher-level performances. Recognizing the need for a more concise instrument, the scientific community seeks a tool that can be easily integrated with other research scales, while practitioners prefer a shorter measure for frequent and convenient assessment. To address this, the present study introduces an 11-item measure of team flow called the Team Flow Quick Scan (TFQS). In addition to validating this measure against the established TFM, this study seeks to replicate the established findings of team flow. While the TFQS measured team flow using a slightly different model from the TFM, the results are consistent with theory and the slightly restricted ability of an 11-item measure to have the sensitivity of a 37-item measure. Ultimately, results replicated the efficacy of the TFM and introduced the TFQS as a viable tool for measuring flow in teams, which can be a powerful aid in helping teams hit their highest performance levels.
Personalization in the workplace may be used to support workers by simplifying tasks or reducing workload, but requires the collection of personal data, and this raises concerns over privacy and possible discrimination if applied indiscriminately. The study explored the factors associated with acceptance of three types of workplace personalization system (Suggester, Swappers and Controller, within-subjects) and two personal data types (Heart Rate, Performance, between-subjects) by presenting vignettes using an online experimental platform (Prolific.com) and capturing respondents' (n = 204) attitudes using recognized acceptance questionnaires (e.g., the Advanced Transport Telematics Acceptance Assessment, [ATTAA]). Results show acceptance is influenced by the type of personalization system, particularly when physiological (heart rate) data is used, with "Swapper" systems receiving the highest ratings for "Usefulness" and "Satisfying" (interpreted as higher acceptance) compared to suggesters and controllers. Acceptance ratings were not significantly different between personalization types when performance data were used. The Affinity for Technology (AFT) and Need for Cognition (NFC) scales were used to categorize participant characteristics, but only revealed significant differences associated with NFC and usefulness-most notably when using performance data. Overall, the results support the need to consider the type of intervention and the type/amount of personal data required when designing and implementing workplace personalization systems and highlight a particular need for caution when physiological data is required.
Intelligent technologies have shifted operator tasks from physical to cognitive demands, making it crucial to understand the impact of task load on visual search tasks. This study examined the effects of auditory working memory and visual perceptual load on inspection tasks using electroencephalography, electrooculography, and electromyography (EMG). Working memory load was varied by an N-back task, and perceptual load by interference presence. Results indicated: (1) Excessive auditory working memory load induces interference suppression, with event-related potential P300 and P200 amplitudes indicating sensitivity to perceptual load; (2) Delta, theta, alpha, and low beta band power spectral densities (PSDs) are sensitive to working memory load, with alpha being most sensitive; (3) Eye blink rate (EBR) increases with working memory load; (4) EMG activity detection may not effectively detect working memory load despite observed trends; (5) Strong positive correlations exist within PSD bands and between P300/P200 amplitudes at certain electrodes, with significant correlations between different modality indicators, such as negative correlations between EBR and low beta band PSD and positive correlations between root mean square, integrated EMG (iEMG), and F4 electrode P300 amplitude.
Collisions between mobile machines and pedestrians persist across worksites despite widespread use of camera monitor systems (CMS) that satisfy current design requirements. This signalizes a potential specification-to-reality gap and a lack of human-centered, field-relevant design criteria for visibility aids in mobile machinery. This study addresses this gap using cognitive work analysis to model generic mobile machine operations in construction sites. Focusing on work domain analysis (WDA), an abstraction hierarchy was developed from document reviews, field observations, and subject-matter expert interviews (n = 5), and validated in two workshops (n = 12). The abstraction hierarchy reveals dense means-ends coupling: CMS effectiveness is highly sensitive to real-world degradations, operator workload, and task dynamics. Based on this model, we clarify the constraints shaping near-field pedestrian detection within the machine's hazard envelope and translate those constraints into a taxonomy of risk factors aligned with ISO 6385 work-system design principles. By making these constraints explicit and measurable, our approach identifies leverage points that yield actionable recommendations for improving CMS design and use. This work enables future hypothesis-driven experiments under representative conditions, supports targeted design interventions and training, and motivates updates to standards and test practices so that CMS performance better reflects field realities to promote safety.
Sports bras function as a form of personal protective equipment, designed to minimize breast motion and pain. The sports bra design significantly affects breast support and discomfort. Although the use of underwire has been shown to enhance bra support, no quantitative data related to the relationship between underwire properties and bra performance has been published to guide designers and manufacturers in selecting underwire. This study aimed to quantitatively examine the relationship between underwire rigidity and breast motion and discomfort. Ten participants were recruited to perform treadmill running under six bra conditions: one without underwire and five with underwires of varying levels of rigidity. A decreasing trend was observed in vertical breast displacement, perceived breast movement, and discomfort as underwire rigidity increased. Significant exponential relationships were identified between underwire rigidity and vertical breast displacement. However, no significant associations were found for perceived breast movement and perceived breast discomfort. These findings underscore the importance of selecting appropriately rigid underwires for sports bras, under the tested conditions (young women with bra size 75 C performing treadmill jogging/running at 7.5 and 10 km/h using the same bra model), offering valuable insights into the design and manufacturing of sports bras in the underwear industry.
The objectives of this work were to investigate the influence of sex, handedness, and lift origin and their interactions on peak lumbar moments and loads at the L5-S1 disc level during maximally loaded box lifting tasks, examine how different normalization techniques can affect these results, and perform a risk assessment of maximum lifting where individuals self-select a safe maximum weight to lift. Twenty participants (sex, 10 women and 10 men; handedness, 10 left-dominant and 10 right-dominant) performed individualized maximally weighted box lifts from five lift origins. The OpenSim Lifting Full-Body Model was used to predict peak lumbar moments and loads. Box lift origin had a consistent influence on peak lumbar moments and loads at the L5-S1 disc level during maximally loaded box lifting tasks, where symmetric lifts produced higher lumbar flexion moments, compression loads, and shear loads than asymmetric lifts. Handedness did not have a consistent influence on peak lumbar moments or loads. Sex consistently influenced lumbar loading, with men experiencing higher lumbar loads than women. It was determined that normalization techniques can alter statistical findings, which can lead to misinterpretations if the approach does not ensure a direct comparison between groups. The risk assessment of maximum lifting revealed that a large percentage of the lifts exceeded the recommended thresholds established in the literature, indicating that individuals are unlikely to self-select a safe maximum weight to lift.
With the increasing application of robots in industrial manufacturing, ensuring the safety of robots designed for close-proximity human-robot collaboration (HRC) has become paramount. Trust is a critical factor influencing human-perceived safety and is essential for sustainable HRC. However, the integration of trust-related influences into safety strategies for close-proximity HRC remains underexplored. This study aims to investigate safety control mechanisms centered around human trust in robots within close-proximity HRC environments. Task performance and safety performance are considered as key factors affecting trust, and Bayesian inference is employed for human trust assessment. Additionally, the concept of a danger field (DF) is introduced, and the DF calculation formula is refined to enhance robot danger assessment. Building on trust and danger assessments, a safety control strategy is designed by integrating the relationship between trust and safety. To validate the effectiveness of this strategy, a scenario is constructed in the Virtual Robot Experiment Platform where a digital human and a UR5 robot collaboratively transport goods, and the safety system is subsequently integrated to conduct simulation experiments. The results indicate that the implementation of the safety control strategy leads to a reduction in peak robot joint speed, a significant decrease in the maximum, mean, and variance of danger assessment values, and a rapid response capability to shut down the system under high-risk conditions. These findings confirm the reliability of the proposed safety control framework.
Safe surgical practice depends on both technical expertise and strong non-technical skills. Scrub nurses play a key role in the operating room, yet their performance may be affected by high mental workload. This cross-sectional study evaluated non-technical skills and mental workload among 200 scrub nurses in five educational hospitals in 2025. Data were collected using the NASA-TLX and SPLINTS tools. Results showed that while scrub nurses had generally good non-technical skills, they also experienced high mental workload. Among female nurses, higher frustration levels were linked to lower performance in assertiveness and information recognition. In nurses with temporary employment, greater mental demand correlated positively with some skill components. Older, more experienced nurses scored lower in information gathering. As the scores of mental workload domains are negatively correlated with scrub nurses' non-technical skills domains, targeted strategies to reduce mental workload may improve non-technical skills performance. Further research is recommended for exploring these results.
Standard operating procedures (SOPs) are widely recognized as essential in maintaining safe operations in high-risk industries, such as oil and gas and petrochemicals. However, limited research has been conducted on the discrepancies between the intended work process (Work as Imagined or WAI) and the actual work process (Work as Done or WAD) under normal working conditions, particularly in these industries. While employees may not always strictly adhere to procedure steps when executing tasks, designing SOPs that allow for adaptation to changing conditions while maintaining adherence remains a challenge. To address this gap, a new approach is proposed in this study that combines two concepts: Hierarchical Task Analysis (HTA) and Abstraction Hierarchy (AH) from work domain analysis. HTA breaks down tasks into a hierarchy of subtasks, while AH decomposes the procedural system into four levels with means-ends relationships. The combination forms the Performance Analysis Technique (PAT). The PAT approach is demonstrated using an SOP for a column flushing task that is part of a three-phase separation system. The results showed that the PAT could visually demonstrate where and how workers make deviations and adaptations to complete a task. This new approach has the potential to improve the design and implementation of SOPs in high-risk industries, enhancing safety and operational efficiency in these environments. The study also highlights the importance of collaboration between procedure writers and frontline workers to design more flexible procedures that recognize adaptation risks.
This study aims to examine how digital addiction based experiences of employees influence attention processes, safety awareness, and collective responsibility in the workplace. It particularly seeks to identify challenges in digitally intensive environments regarding cognitive load, behavioral reflexes, and institutional risk communication. The research employed a phenomenological design with semi- structured interviews conducted with 19 employees from various sectors across Turkey. Criterion sampling was used, and interviews continued until data saturation was reached. All interviews were audio recorded, transcribed verbatim, and analyzed using Colaizzi's (1978) seven step method. Reporting followed the COREQ checklist to ensure rigor and transparency. Thematic analysis revealed three main themes: Cognitive Load and Deterioration in Safety Perception, Behavioral Disconnection and Erosion in Safety Attitudes, and Organizational Silence and Collective Collapse in Risk Awareness. These findings show that digital addiction negatively affects attention, reflexes, safety behaviors, and institutional reporting processes. Digitalization influences not only work efficiency but also employees' safety behaviors and organizational risk communication. Constant digital stimuli cause distraction, disrupt safety practices, and weaken collective responsibility. The study provides insights into how digital addiction transforms workplace safety culture and recommends comprehensive prevention strategies for employers, managers, and policymakers.
In complex sociotechnical systems, situational awareness (SA) emerges not from individuals alone but from the shared understanding distributed across human agents, automated subsystems, and information artefacts. Effective coordination therefore depends on distributed situational awareness (DSA), with human-machine interfaces (HMIs) playing a central role. However, traditional HMI evaluations often overlook how DSA is structured by agents' roles, task interdependencies, and communication networks. This paper introduces a method that extends the Event Analysis of Systemic Teamwork (EAST) with an entropy measure, derived from Hick's Law, to quantify informational complexity. The integration provides a systematic means of assessing whether information distribution supports task demands and agent capacity. A walkthrough application illustrates how the method detects DSA misalignments and guides reconfiguration by redistributing information to balance entropy. Results show that entropy-enhanced EAST highlights critical nodes and offers deeper insight into DSA dynamics. The findings also reveal how interface design choices -such as deliberately increasing complexity to enforce communication-shape the distribution of awareness across agents.
Considering the COVID-19 pandemic, understanding the behavior of social groups on crosswalks has become increasingly important. This field study aims to investigate the dynamic behavior of pedestrian groups crossing streets during the pandemic. By analyzing the movement patterns and behaviors of group under different conditions, we aim to provide insights into how crosswalk infrastructures can be improved for a safer and more efficient pedestrian experience during the pandemic. An observation experiment was conducted at a crosswalk near Shinjuku Station in Japan in 2021. Trajectories of 296 groups were analyzed with up to five members, examining their velocity characteristics and spatial relations. Our findings reveal several important insights. First, the group size affects the average speed and speed difference of group members, and that the width of the crosswalk is positively correlated with the average speed of group members. Second, the group size is uncorrelated with mean inter-distance of members, regardless of group size. However, the farther members are from the leader, the greater the average distance between adjacent members. Third, distinct group shapes for groups of different sizes. Groups of three people form a V shape, while groups of four people form a U shape, and groups of five people form a trapezoid shape. Finally, the group size is correlated with the average offset angle of group members in the northbound scenario. Larger groups tend to have smaller average offset angles, making them more likely to choose the shortest route. Overall, this study is a crucial step in developing safety-oriented walking modeling tools for pedestrians at intersections. It also has important implications for predicting pedestrian crossing behavior and show how human factors and ergonomics methods can be used to improve urban public infrastructure systems in densely populated countries, such as Japan.
Despite comprising 18% of the Australian population, older adults account for 40% of pedestrian fatalities. It has been proposed that age-related decline in perceptual, cognitive, and physical function contributes to these deaths. To date, the important safe street-crossing skills of hazard perception and gap acceptance have been understudied in an older population and would benefit from being examined using immersive technologies, such as virtual reality (VR). Using a mixed-method design and adopting human factors and ergonomics principles, this study determined the feasibility and acceptability of a protocol using a VR pedestrian street-crossing task (VR-PSCT), including the presence of cybersickness. Data were collected from 14 younger adults (25-45 years) and 14 older adults (> 60 years). Participants completed tasks that measured visual perceptual capacity (e.g., visual acuity), cognitive capacity (e.g., visuospatial attention), and physical capacity (e.g., balance). Hazard perception and gap acceptance were measured using a VR headset where a series of 360 degrees video clips captured from real-world pedestrian situations were presented. Hazard perception response time did not differ between older and younger adults, nor did their hazard perception accuracy scores; however, gap acceptance response time was significantly slower for older adults compared with younger adults. The older adults reported that the protocol length was too long and induced high levels of fatigue. The VR-PSCT was well tolerated, with some instances of mild cybersickness and motor instability for the older adults. This study has established the feasibility of our VR-PSCT task and protocol and highlighted several user-centered modifications needed to conduct further testing with a larger cohort of older adults. By using the latest immersive technologies, we can obtain a greater understanding of older adult pedestrian behaviors and the factors that predict these behaviors.
