Human Factors and Ergonomics in Manufacturing & Service Industries最新文献

Human error is often implicated in industrial accidents and is frequently found to be a symptom of broader issues within the sociotechnical system. Therefore, research exploring human error during maintenance activities is important. This article aims to assess the probability of human error in maintenance tasks at a cement factory using the Cognitive Reliability and Error Analysis Method and System Dynamics modeling. Given that human error probability (HEP) is influenced by various common performance conditions (CPCs) and their sub-factors, and changes dynamically in response to other variables, the SD method offers a practical approach for estimating and predicting human error behavior over time. This study identifies and quantifies the variables affecting HEP, explores their interactions and feedback in maintenance tasks, and assesses the associated costs. The machine learning technique is then used to estimate the relationship between HEP and these costs. The optimal value of the HEP function, 0.000772, is determined by identifying the minimum point of a cubic function, thereby minimizing associated costs and occupational accidents. Determining the optimal HEP is crucial for minimizing excessive costs and investing in improved ergonomics and CPCs for better performance. This addresses a significant gap in existing research where the impact of human error on maintenance tasks has not been estimated as a function. Furthermore, three scenarios are presented to help managers allocate the organization's budget more effectively.

The recent advancement in additive manufacturing (AM) leads to an extensive need for an industrial workforce in the near future. Workforce training in AM requires expensive capital investment for installing and maintaining this technology and proper knowledge about potential safety hazards. Traditional classroom settings often fail to bridge the critical gap between textbook learning and practical applications. Virtual reality (VR) training can simulate real-world scenarios in a safe and controlled environment and improve student involvement to foster practical learning. In this paper, a virtual training platform for 3D printing has been developed and studied to improve AM education. The developed environment contains a selective laser sintering printer, a preparation station with necessary supplies, a control panel for process planning, and a post-processing station. This platform provides students with excellent learning opportunities to gain hands-on experiences and critical engineering skills on operating process parameters and safety measures. Undergraduate students majoring in industrial engineering were exposed to this learning approach to enhance their engagement and cognitive processing skills. Students' attentions were measured using eye metrics (fixation duration and preference index), and their exposure experiences were collected through the simulation sickness questionnaire, presence questionnaire, and system usability scale. Pre- and post-VR training questionnaires and performance metrics (task completion time and accuracy) evaluated students' learning outcomes. Results provide valuable insights into students' attention, performance, and satisfaction with virtual training environments. Users' gaze behavior and subjective responses revealed many challenges that will help future researchers develop assistive instructions within this virtual educational platform.

The workload levels of pilots directly affect their flight performance and the safety of the whole flight. To explore the real-time workload of pilots in different flight phases (takeoff, cruise, and landing), this paper leveraged National Aeronautics and Space Administration Task Load Index (NASA-TLX), a subjective evaluation scale, and PhotoPlethysmoGraphy (PPG) signals of 21 participants using a flight simulator and a wearable sensor. First, the workloads of pilots under different phases were explored by the NASA-TLX scales; secondly, the pulse rate variability (PRV) features were selected by variance analysis and random forest importance evaluation; finally, the performances of the k-nearest neighbor (KNN), random forest (RF), and support vector machine (SVM) were compared for workload levels identification. It is shown that the workloads are ranked as follows: landing > takeoff > cruise. SDNN, CVCD, CVNNI, LF, TP, SD2, and SD2/SD1 were used as selected features with significant differences in different flight phases. In addition, machine learning models can effectively identify pilot workloads, and feature selection enhances the performance of both KNN and RF classifiers. The best identification of workload was achieved using the selected PRV features as inputs to the KNN classifier, with an average accuracy of 88.9%. Our results indicate that the KNN classifier and PRV features are suitable for identifying pilot workload. The pilot workload is highest during the landing phase, which provides a reference for flight safety management. The findings from this research could contribute to developing a robust pilot workload detection system and improve current flight operation safety regulations.

Every year, many people lose their lives because of road accidents. It is evident from statistics that drowsiness is one of the main causes of a large number of car accidents. In our research, we wish to solve this major problem by measuring the drowsiness level of the human brain while driving. The study aims to develop a novel technique to detect different alertness levels (i.e., awake, moderately drowsy, and maximally drowsy) of a person while driving. A hybrid model using a stacked autoencoder and hyperbolic tangent Long Short-Term Memory (TLSTM) network with attention mechanism is designed for this purpose. The designed model uses different biopotential signals, such as electroencephalography (EEG), facial electromyography (EMG), and different biomarkers, such as pulse rate, respiration rate galvanic skin response, and head movement to detect a person's alertness level. Here, the stacked autoencoder model is used for automated feature extraction. TLSTM is used to predict a person's alertness level using stacked autoencoder network-extracted features. The proposed model can classify awake, moderately drowsy, and maximally drowsy states of a person with accuracies of 99%, 98.3%, and 98.6%, respectively. The novel contributions of the paper includes (i) incorporation of an attention mechanism into the TLSTM network of the proposed hybrid model to focus on the emphatic states to enhance classification accuracy, and (ii) utilization of EEG, facial EMG, pulse rate, respiration rate, galvanic skin reaction, and head movement pattern to assess a person's alertness level.

In the digital interface of multimodal audio–visual presentation, the appearance of irrelevant information often brings cognitive interference or even confusion, leading to decision-making errors when users focus on or manipulate the interface target. However, few studies have explored the brain's inhibition effect and cognitive law evoked by audio–visual interference from the perspective of interface information design. On the basis of Stroop's classic interference task, an experimental paradigm of multimodal audio–visual stimuli to induce event-related potential (ERP) components was designed for digital interfaces in this study. Combining behavioral measurement and ERP technology, this study discussed the differences in the induced inhibition effects between the two carriers under various audio–visual interferences. The findings demonstrated that all five interference stimuli, based on functional icons and Chinese characters, elicited significant N250 and N400, with a similar time course. Compared with the Chinese character group, the functional icon group elicited more negative activity in the frontal and some parietal-occipital regions, indicating that the functional icon required more cognitive inhibitory resources to resist interference stimuli. Moreover, the inhibition effect induced by audio–visual interference with the same semantics was significantly lower than that of opposite semantics and even lower than that of single-sensory interference. The findings offered physiological evidence for the inhibition effect induced by audio–visual semantic interference in digital interfaces and proposed design principles for the interface information of human–machine systems.

Shoe manufacturing companies often use overtime work but neglect the impacts and importance of physical recovery time. Ergonomic methods aim to analyze this, but they focus on deterministic aspects, which limits their ability to evaluate working conditions amid variations over time. This research explores how a simulation model can mitigate these limitations and enhance analysis of overtime and physical recovery on worker absenteeism. The objective was developed a simulation model using System Dynamics (SD) to represent working conditions and assess the influence of overtime and recovery time in Brazil's footwear industry. An Ergonomic Analysis of Work was conducted in a large company's production cell. Using SD, were constructed a causal and simulation model to analyze three scenarios. An additional hour of work increased physical overload by 44%, leading to 5, 4 leave requests, and 48 days of absenteeism per year. Increasing recovery time by 15 min reduced overload to 38,96%, resulting in 4, 9 leave requests and 13,68 days of absenteeism. The SD simulation model mitigated the limitations of ergonomic methods in understanding the dynamic relationships over time, emphasizing the importance of actively managing overtime and physical recovery time.

Due to the physical nature of their work, sonographers are exposed to many musculoskeletal disorder risk factors, including awkward posture, repetitive movements, forceful manual exertion, and static muscle contractions, especially in the upper limbs. The current study is an investigation of musculoskeletal disorders among sonographers, caused by various occupational risk factors via different sonographic scan types. During the first phase of this study, the musculoskeletal symptoms and work postures of 29 subjects were investigated. During the second phase, muscle activity was quantified, and grip/push forces were estimated using the data obtained from 10 volunteer sonographers. 82% of sonographers experienced musculoskeletal symptoms. Based on the final scores and action levels obtained via rapid upper limb assessment, while performing scans of left regions; ergonomic changes and interventions were found necessary, to relieve stress on the sonographer's body. The results of muscular activity per muscle and scan type, showed that the mean muscle activity of the middle deltoid muscle was significantly higher during the right abdominal scan (17.64% maximum voluntary contraction [MVC]), compared to those of thyroid (12.54% MVC) and left abdominal (7.32% MVC) scans. Additionally, mean grip and push forces during both abdominal scans were significantly higher than those during the thyroid scan. Despite an injury risk during all scans, risk factor impact was different among scan types. This groundbreaking study represents the first that captures and measures both grip and push forces simultaneously, which may prove helpful while investigating corrective interventions or optimizing design of sonography robots and ergonomic probes in future studies.

Driver distraction is intricately linked to human behavior and cognitive ergonomics, as it explores how human engagement with various stimuli influences attention and decision-making processes while driving. The main purpose of this study is to comprehensively explore whether using Human–Machine Interface infotainment systems in automated vehicles can affect driver distraction. To this end, driver distraction was measured by driving performance features (speed, lane position, and reaction time), behavioral features (fixation time and pupil dilation), physiological features (changes in oxyhemoglobin), and subjective assessment (NASA-TLX workload). Twenty-one participants equipped with an eye tracker and functional near-infrared spectroscopy drove a driving simulator in the current investigation. The results revealed that interacting with the infotainment systems significantly affects the drivers' average speed (F_2,40 = 13.60, p < .0001), reaction time (F_2,40 = 4.74, p = .0142), fixation time (F_2,40 = 88.61, p < .0001), pupil dilation (F_2,28 = 3.63, p = .0356), and workload (F_2,40 = 14.40, p < .0001). Moreover, driving mode significantly affects drivers' speed deviation (F_2,40 = 6.12, p = .0048), standard deviation of lane position (F_2,40 = 10.57, p = .0002), fixation time (F_2,40 = 36.71, p < .0001), and workload (F_2,40 = 28.08, p < .0001). Drawing from the findings of this article and emphasizing human-centric design principles, researchers and engineers can craft automotive technologies that are intuitive, effective, and safer. This is vital for mitigating driver distraction and guaranteeing the beneficial influence of automated vehicles on both road safety and the overall driving experience.

Functional resonance analysis method (FRAM) is extensively employed in analyzing and managing performance variabilities. Additionally, semi-quantitative and quantitative methods have been increasingly integrated with the FRAM to analyze complex socio-technical systems to improve safety levels. This review article presents a comprehensive and updated survey of current literature focused on semi-quantitative and quantitative methods employed for quantifying performance variabilities and exploring aggregation/propagation rules. A total of 1659 studies published between 2012 and March 2024 from various scientific databases were systematically examined using preferred reporting items for systematic review and meta-analysis, identifying 29 studies that met inclusion criteria. The identified studies were categorized into four groups based on the quantitative methods employed: Monte Carlo simulation, fuzzy logic, cognitive reliability and error analysis method, and miscellaneous approaches. While different methodologies had unique strengths, they commonly relied on expert judgment for data collection, whether for defining probability distributions in Monte Carlo simulations, membership functions, and fuzzy rule bases in fuzzy inference systems, or selecting common performance conditions, determining their interrelationships, and assigning scores. Addressing bias from expert judgment in assessing performance variabilities can be achieved by using suitable experts' opinions integration techniques, and leading safety indicators in the analysis.

An important question in teamwork research is how to maximize performance and the aspects of the team's dynamics and collaboration process that underpin it. Prior research has shown that when team members who are collaborating towards a common purpose experience flow together (team flow; optimal experiences that occur simultaneously at the individual and team levels, entailing deep focus and intrinsic motivation to perform an activity), the team significantly improves its performance and team members experience many positive results at both the individual and team levels. Further advances have built a model of team flow and a means for measuring the construct, as well as qualitative results in business teams to confirm how the elements of team flow interact to generate the positive experiences and higher performance. This study adds practical value to the research by providing proof-of-concept for an intervention that promotes team flow in business teams. This cross-case-study of 15 teams across five different organizations uses the Team Flow Monitor as a barometer of team health and dynamics, which in turn serves as the centerpiece of an iterative intervention protocol for leading/guiding teams in targeted self-reflection that can generate virtuous cycles of improving dynamics and performance. In addition to a significant amount of qualitative data confirming the efficacy of the intervention in enabling teams to overcome obstacles and experience more team flow, quantitative analysis of Team Flow Monitor scores showed an increase on average team flow scores across the teams over the course of the intervention (Cohen's d = 0.6). Implications for translating team flow research to field situations are discussed, along with further potential uses of the Team Flow Monitor.