On-site observations of physical work demands of train conductors and service electricians in the Netherlands

The objective of the present study was to assess the exposure to physical work demands of train conductors and service electricians at a railway company in the Netherlands. On-site observations were performed using the Task Recording and Analysis on Computer observation system to identify the mean duration and frequency of tasks, activities and body postures. In total, 36 train conductors and 41 service electricians were observed for a net working day of 7 h. Results showed that train conductors and service electricians climbed a flight of stairs on average 249 and 258 times, respectively, and that service electricians worked above shoulder height for 65 min on average. In both jobs, guidelines were exceeded, e.g. duration of standing, the number of times climbing a flight of stairs, kneeling and squatting or working in awkward postures, which are risk factors for developing musculoskeletal complaints. Suggestions are made concerning how to reduce these risk factors.

Statement of Relevance: Understanding physical work demands is essential for recognising risk factors for musculoskeletal disorders. Since train conductors and service electricians in the Netherlands reported work-related complaints, on-site observations were performed to determine the duration and frequency of physical work demands. Risk factors were identified in both professions, providing insights concerning preventative measures.     

Convolutional long short-term memory model for recognizing construction workers’ postures from wearable inertial measurement units

This paper proposes using Deep Neural Networks (DNN) models for recognizing construction workers’ postures from motion data captured by wearable Inertial Measurement Units (IMUs) sensors. The recognized awkward postures can be linked to known risks of Musculoskeletal Disorders among workers. Applying conventional Machine Learning (ML)-based models has shown promising results in recognizing workers’ postures. ML models are limited – they reply on heuristic feature engineering when constructing discriminative features for characterizing postures. This makes further improving the model performance regarding recognition accuracy challenging. In this paper, the authors investigate the feasibility of addressing this problem using a DNN model that, through integrating Convolutional Neural Networks (CNN) with Long Short-Term Memory (LSTM) layers, automates feature engineering and sequential pattern detection. The model’s recognition performance was evaluated using datasets collected from four workers on construction sites. The DNN model integrating one convolutional and two LSTM layers resulted in the best performance (measured by F1 Score). The proposed model outperformed baseline CNN and LSTM models suggesting that it leveraged the advantages of the two baseline models for effective feature learning. It improved benchmark ML models’ recognition performance by an average of 11% under personalized modelling. The recognition performance was also improved by 3% when the proposed model was applied to 8 types of postures across three subjects. These results support that the proposed DNN model has a high potential in addressing challenges for improving the recognition performance that was observed when using ML models.     

Gravitational demand on the neck musculature during tablet computer use

Tablet computer use requires substantial head and neck flexion, which is a risk factor for neck pain. The goal of this study was to evaluate the biomechanics of the head–neck system during seated tablet computer use under a variety of conditions. A physiologically relevant variable, gravitational demand (the ratio of gravitational moment due to the weight of the head to maximal muscle moment capacity), was estimated using a musculoskeletal model incorporating subject-specific size and intervertebral postures from radiographs. Gravitational demand in postures adopted during tablet computer use was 3–5 times that of the neutral posture, with the lowest demand when the tablet was in a high propped position. Moreover, the estimated gravitational demand could be correlated to head and neck postural measures (0.48 < R ² < 0.64, p < 0.001). These findings provide quantitative data about mechanical requirements on the neck musculature during tablet computer use and are important for developing ergonomics guidelines.

Practitioner Summary: Flexed head and neck postures occur during tablet computer use and are implicated in neck pain. The mechanical demand on the neck muscles was estimated to increase 3–5 times during seated tablet computer use versus seated neutral posture, with the lowest demand in a high propped tablet position but few differences in other conditions.     

Relation between perceived and measured workload obtained by long-term inclinometry among dentists

Dentists reported high perceived physical work conditions. Working postures and movements of the head and upper extremities during dental work were registered with inclinometry measurements during four hours. The aim was to clarify the relationship between measured working postures/movements and perceived physical work conditions. Dentists worked with elevated arms and a rather steep forward inclination of the head. Correlations (r = −0.52 to -0.66) between inclination velocity and perceived workload on VAS scales were found, but there were only weak correlations between observed working postures. The different tasks involved in dental work provide limited variation in work movements and postures, measured by inclinometry. By alternating between sitting and standing, it might be possible to achieve variation in physical workload during dental work.     

Changes in posture through the use of simple inclines with notebook computers placed on a standard desk

This study evaluated the use of simple inclines as a portable peripheral for improving head and neck postures during notebook computer use on tables in portable environments such as hotel rooms, cafés, and airport lounges. A 3D motion analysis system measured head, neck and right upper extremity postures of 15 participants as they completed a 10 min computer task in six different configurations, all on a fixed height desk: no-incline, 12° incline, 25° incline, no-incline with external mouse, 25° incline with an external mouse, and a commercially available riser with external mouse and keyboard. After completion of the task, subjects rated the configuration for comfort and ease of use and indicated perceived discomfort in several body segments. Compared to the no-incline configuration, use of the 12° incline reduced forward head tilt and neck flexion while increasing wrist extension. The 25° incline further reduced head tilt and neck flexion while further increasing wrist extension. The 25° incline received the lowest comfort and ease of use ratings and the highest perceived discomfort score. For portable, temporary computing environments where internal input devices are used, users may find improved head and neck postures with acceptable wrist extension postures with the utilization of a 12° incline.     

Assessment of workload and arm position during different work sequences: a study with portable devices on construction workers

It is recognised that work related shoulder pain is overrepresented among construction workers compared to other occupations. Studies have shown that working with hands above shoulder level increases the shoulder load. Most studies have been confined to the laboratory. The present project was carried out to map the muscular engagement and postures of construction workers undertaking ceiling fitting, and to compare the results to those of the laboratory studies. Two ambulatory devices were used, one allowing recording of electromyographic (EMG) signals bilaterally from the trapezius muscle, and the other to record the position of both arms and back by means of measuring the angles between the vertical line and the back and both upper arms. These recordings were performed during 1.5–2 h work sequences. The results show that the work was mostly performed in an upright position, that both arms were used to a similar amount and that the workers for a large proportion of their working time had their upper arms at levels that are considered harmful in view of shoulder load. The EMG data showed that nearly 50% of the work was spent with trapezius activity that exceeded that of the reference contraction used (about 15% of maximal voluntary contraction) and that the time spent in muscular relaxation was 10%. It was concluded that the exposure of construction workers undertaking ceiling fitting meets the criteria formulated on the basis of laboratory experiments with respect to a high risk of acquiring chronic shoulder pain, due to rotator cuff tendinitis.     

复杂姿态下的安全帽佩戴检测方法研究

在施工人员复杂姿态下,现有安全帽佩戴检测方法存在检测难度大,精度不高的问题,提出一种基于头部识别的安全帽佩戴检测方法。通过肤色特征识别和头部检测获取头部区域,并进行交叉验证确定头部区域,使用YOLOv4目标检测网络识别安全帽,通过安全帽区域与头部区域的位置关系判断安全帽佩戴情况。最后,通过实验对比分析其他安全帽佩戴检测方法的性能,对安全帽佩戴检测方法进行总结并提出展望。     

Observations of working postures in garages using the Ovako Working posture Analysing System (OWAS) and consequent workload reduction recommendations

The working postures of mechanics (n = 84) in 42 garages were observed using the Ovako Working posture Analysis System (OWAS). During observation, both working postures and work activities were recorded. A computer program was developed for the data analyses. Using this program it is possible to calculate the working posture load for each work activity and the contribution of a specific activity to the total working posture load. This is a substantial extension of the original OWAS method. Five out of 19 observed postures of the body members were classified as Action Category 2, which suggests they were slightly harmful to the musculoskeletal system and likely to cause discomfort. Of the so-called typical working postures, 31.9% was classified in Action Category 2, suggesting that during a substantial part of the working day typical working postures occur which are at least slightly harmful to the musculoskeletal system. Moreover, those work activities principally causing the workload to fall in OWAS' higher Action Categories were identified. For each of these three work activities an alternative work method was observed. The data show that in all three work activities the use of a vehicle lift reduces the number of poor working postures thereby reducing the load on the musculoskeletal system.     

Simultaneous measurements of posture and movements of head and trunk by continuous three-dimensional registration

The postures of three groups of employees were measured: straddle carrier drivers, crane operators and office employees. This type of sedentary work can be characterized as being highly static. Using a continuous three-dimensional registration device, the postures and movements of head and trunk were recorded simultaneously. The results show that the adopted postures and patterns of movement were predominantly imposed by the workplace. The posture of the crane operators was the most static compared to the other occupations. The most adopted posture in the sagittal plane for crane operators was trunk flexion of 5° and head flexion of 60°. Typically, a straddle carrier driver rotated his head more than 45° to the left or right for 28% of the day, which far exceeded that of the other groups. The measuring device provides accurate and reproducible data that can subsequently be used for calculating the postural load and for ergonomic analysis.     

Stabilisation times after transitions to standing from different working postures

Transitioning to standing after maintaining working postures may result in imbalance and could elicit a fall. The objective of this study was to quantify the magnitude of imbalance using a stabilisation time metric. Forty-five male participants completed three replications of conditions created by one of four working postures (bent at waist, squat, forward kneel, reclined kneel) and three durations within posture. Participants transitioned to quiet standing at a self-selected pace. Stabilisation time, based on changes in centre of pressure velocity, was used to indicate the initiation of steady state while standing. Stabilisation time was significantly affected by static postures but not duration within posture. The largest stabilisation times resulted from transitions initiated from a bent at waist posture. The smallest were associated with the kneeling postures, which were not significantly different from each other. Findings may lead to recommendations for redesign of tasks, particularly in high-risk environments such as construction.

Statement of Relevance: Task performance on the jobsite often requires individuals to maintain non-erect postures. This study suggests that working posture affects stabilisation during transition to a standing position. Bending at the waist and squatting resulted in longer stabilisation times, whereas both kneeling postures evaluated resulted in greater imbalance but for a shorter duration.     

Applying incremental Deep Neural Networks-based posture recognition model for ergonomics risk assessment in construction

Monitoring and assessing awkward postures is a proactive approach for Musculoskeletal Disorders (MSDs) prevention in construction. Machine Learning models have shown promising results when used in recognition of workers’ posture from Wearable Sensors. However, there is a need to further investigate: i) how to enable Incremental Learning, where trained recognition models continuously learn new postures from incoming subjects while controlling the forgetting of learned postures; ii) the validity of ergonomics risk assessment with recognized postures. The research discussed in this paper seeks to address this need through an adaptive posture recognition model– the incremental Convolutional Long Short-Term Memory (CLN) model. The paper discusses the methodology used to develop and validate this model’s use as an effective Incremental Learning strategy. The evaluation was based on real construction workers’ natural postures during their daily tasks. The CLN model with “shallow” (up to two) convolutional layers achieved high recognition performance (Macro F1 Score) under personalized (0.87) and generalized (0.84) modeling. Generalized CLN model, with one convolutional layer, using the “Many-to-One” Incremental Learning scheme can potentially balance the performance of adaptation and controlling forgetting. Applying the ergonomics rules on recognized and ground truth postures yielded comparable risk assessment results. These findings support that the proposed incremental Deep Neural Networks model has a high potential for adaptive posture recognition. They can be deployed alongside ergonomics rules for effective MSDs risk assessment.     

The ratio of thoracic to lumbar compression force is posture dependent

Despite the evidence suggesting that between 8% and 55% of manual labourers experience thoracic pain, research on spinal loading during occupational tasks has been almost invariably limited to the lumbar spine. In this study, we determined the ratio of thoracic to lumbar compression force and the relative risk of injury to each region in various postures. Compressive forces on the spine were calculated based on previously reported thoracic and lumbar intradiscal pressures and disc cross-sectional areas. Flexion postures were associated with an approximate doubling in lumbar compression force but only small increases (or even decreases) in thoracic compression. The ratio of thoracic to lumbar compression was above the tolerance ratio (i.e. the ratio of thoracic to lumbar compressive strength) during upright postures and below the tolerance ratio during flexion postures, indicating that upright postures may pose a greater relative risk of injury to the thoracic spine than to the lumbar spine.

Practitioner summary: Previously reported thoracic and lumbar in vivo disc pressures during various postures were compared. The ratio of thoracic and lumbar compression increased during upright postures and decreased in flexed postures, indicating that upright postures may pose a greater risk of injury to the thoracic spine than to the lumbar spine.     

Study on the variation of peak isometric strength and EMG activity in static field-simulated lifting postures

Peak isometric strength was measured from 10 adult Indian construction workers in eight different field-simulated (FS) postures. This peak-strength data in these FS postures were compared with symmetric postures. In symmetric postures, the vertical load positions were kept the same as FS postures and the points of force exertion were fixed at 40 cm distance in front of the subject. From both symmetric and FS static strength data, it was shown that the maximum peak strength occurred at medium vertical height level and decreased with both increase and decrease of the vertical height level. The maximum and minimum peak strengths were obtained in different FS postures as 222.85±61.15 N and 85.65±19.89 N, respectively. It was observed that the lifted weight in the field was 12.0 kg, which corresponds to 54.54% and 137.3% (i.e. >100%) of these maximum and minimum peak-strength values. This result indicated the prevalence of high risk factors in the field. During this study, surface EMGs from four different muscles (i.e. trapezius, external oblique, rectus abdominis and erector spinae) were collected while exerting the peak isometric strength. From ANOVA analysis, it was shown that the erector spinae and trapezius activities were significantly (p⩽0.05) related to the peak-strength value, whereas external oblique and rectus abdominis activities were not. It was also observed that RMS of erector spinae activity increases with ipsi-lateral increase of asymmetry angle along with the decrease in maximum static peak-strength level in FS postures.Relevance to industryThe peak static strength and EMG activities were measured in FS postures to highlight the potential risk factors in building construction industry. This study was conducted on adult female building construction workers, where the female construction workers were not well represented in the literature. Moreover, from this study, it is very clear that the actual field postures are of more complex in nature than the simulation studies, as mentioned in earlier studies.     

Analysis of working postures in hammering tasks on building construction sites using the computerized OWAS method

The main objectives of this study were to identify the most problematic postures in hammering tasks performed at building construction sites through application of the computerized OWAS method, and to develop recommendations for improvement of working method and workplaces. Eighteen construction workers, with mean age of 41.6, from three construction companies participated in the field study. The hammering tasks observed during the two-month period included roof boarding, concrete form preparation, clamping support braces, assembling roof frames, roof joisting, shelter form preparation, and fixing fork clamps. Three different types of hammer, including a small Fiskar's hammer, a Fiskar's construction hammer, and a Rocket hammer, were used by the workers. Of all the observations, poor working postures were observed most frequently in roof joisting (12.4% of all observations within the task), followed by concrete form preparation (8.6%), and construction of frames for the roof (7.5%). Overall, out of 593 different postures analysed, a total of 7.8% of postures adopted by the workers during various hammering tasks were classified into OWAS categories III or IV, indicating that these postures should be corrected either soon or immediately. The computerized OWAS method for postural data analysis proved to be a very useful way to reduce postural load of dynamic hammering tasks, and allowed for efficient application of the original OWAS method.     

Investigation of optimal table height and surface angle in meatcutting

Butchers have a high frequency of occupational disorders or pain localised to their hands, shoulders, low back and neck, compared with other occupational groups. It is likely that the disorders are caused by high loadings of these parts of the body during work. The high loadings have been found to be caused by high resistance when cutting, inappropriate working postures and by manual handling of material. To improve the butchers' working conditions by providing means for appropriate working postures, cutting tables with adjustable height and surface angle were constructed and tested. Evaluations were made during work in the laboratory and in the industry. The aim of these tests was to find the optimal height of the table for each butcher and to find out if a tilted surface would improve the working postures and be accepted by the butchers. The evaluations were made by biomechanical calculations from photos, ergonomic analysis of videotapes taken during the tests and by using ranking scales. The results show that (1) one optimal height of the table was not to be found, meaning a height resulting in a posture that causes no load on both back and shoulders at the same time; (2) a table height of 17 to 22 cm below elbow height resulted in low loadings on both the low back and the shoulders and was ranked as the most comfortable height; (3) tilting of the surface 5 degrees to 10 degrees was accepted by the butchers and ergonomically favourable; (4) the possibility of altering the table height during the workday was perceived as unloading since the load could be shifted between different parts of the body.     

Endurance, pain and resumption in fully flexed postures

In this study effects of low loads in fully flexed postures were investigated. Thirteen men who were unused to the postures participated. Thirteen professional construction workers with long experience of suchlike postures were also studied. Pain reactions during and after loading were observed, as well as endurance time and the recovery process, here by studying the resumption time. Endurance and resumption times differed little from those given by models used for more common postures. Pain from the legs and not from the back limited the working ability in 86% of the endurance tests. Thirdly, the construction workers had significantly longer endurance time and shorter resumption time.     

An OWAS-based analysis of nurses' working postures

Abstract

The working postures of Dutch nurses (n = 18) in an orthopaedic ward and a urology ward were observed using-the Ovako Working posture Analysis System (OWAS). During observation, both working postures and activities were recorded. A specially developed computer program was used for data analysis. By means of this program, it was possible to calculate the working posture load for each activity and the contribution of a specific activity to the total working posture load. This study shows that some activities of the nurses in both wards were performed with poor working postures. In the orthopaedic (resp. urology) ward two (resp. one) out of 19 observed postures of parts of the body were classified as Action Category 2. Moreover, 20% (resp. 16%) of the so-called typical working postures was classified in Action Category 2. This suggests, that in both wards working postures that are slightly harmful to the musculoskeletal system, occur during a substantial part of the working day. Differences between both wards with respect to working posture load and time expenditure were determined. Activities causing the workload to fall into OWAS higher Action Categories were identified. The data show that poor working postures in the nursing profession not only occur during patient handling activities but also during tasks like ‘administration’. Focusing on patient-handling (i.e., lifting patients) in order to determine the load on the musculoskeletal system would therefore lead to an underestimation of the total working posture load of nurses     

Electromyography of the thigh muscles during lifting tasks in kneeling and squatting postures

Underground coal miners who work in low-seam mines frequently handle materials in kneeling or squatting postures. To assess quadriceps and hamstring muscle demands in these postures, nine participants performed lateral load transfers in kneeling and squatting postures, during which electromyographic (EMG) data were collected. EMG activity was obtained at five points throughout the transfer for three quadriceps muscles and two hamstring muscles from each thigh. ANOVA results indicated that EMG data for nine of 10 thigh muscles were affected by an interaction between posture and angular position of the load lifted (p < 0.001). Muscles of the right thigh were most active during the lifting portion of the task (lifting a block from the participant's right) and activity decreased as the block was transferred to the left. Left thigh muscles showed the opposite pattern. EMG activity for the majority of thigh muscles was affected by the size of the base of support provided by different postures, with lower EMG activity observed with a larger base of support and increased activity in postures where base of support was reduced (p < 0.05). Thigh EMG activity was lowest in postures with fully flexed knees, which may explain worker preference for this posture. However, such postures are also associated with increased risk of meniscal damage. STATEMENT OF RELEVANCE: Kneeling and squatting postures are sometimes used for manual lifting activities, but are associated with increased knee injury risk. This paper examines the EMG responses of knee extensors/flexors to lifting in these postures, discusses the impact of posture and kneepads on muscle recruitment and explores the implications for work in such postures.     

Notebook computer use on a desk,lap and lap support: Effects on posture,performance and comfort

This study quantified postures of users working on a notebook computer situated in their lap and tested the effect of using a device designed to increase the height of the notebook when placed on the lap. A motion analysis system measured head, neck and upper extremity postures of 15 adults as they worked on a notebook computer placed on a desk (DESK), the lap (LAP) and a commercially available lapdesk (LAPDESK). Compared with the DESK, the LAP increased downwards head tilt 6° and wrist extension 8°. Shoulder flexion and ulnar deviation decreased 13° and 9°, respectively. Compared with the LAP, the LAPDESK decreased downwards head tilt 4°, neck flexion 2°, and wrist extension 9°. Users reported less discomfort and difficulty in the DESK configuration. Use of the lapdesk improved postures compared with the lap; however, all configurations resulted in high values of wrist extension, wrist deviation and downwards head tilt.

Statement of Relevance: This study quantifies postures of users working with a notebook computer in typical portable configurations. A better understanding of the postures assumed during notebook computer use can improve usage guidelines to reduce the risk of musculoskeletal injuries     

基于自适应线性回归的头部姿态计算

减少训练样本数量和不使用任何硬件参数是目前头部姿态计算领域的最大挑战。针对这些挑战,提出一种仅需要54个训练样本的头部姿态估计方法。其基本思想是将关键点间的归一化距离作为输入特征向量,并利用一阶范数最小化稀疏地挑选出一组图片集,这些图片对应头部姿态的线性组合为测试图片的头部姿态。实验表明该方法在不知道任何硬件参数的条件下,可以实现单方向上3?的头部姿态估计。此外,该方法也能适用于不均匀光照条件和较大的头部转动情况,且计算精度高于与之相关方法。