International Journal of Industrial Ergonomics 67 (2018) 41–52

Contents lists available at ScienceDirect

International Journal of Industrial Ergonomics

journal homepage: [Link]/locate/ergon

Biomechanical exposure of industrial workers – Influence of automation T

process
Francisco Locksa, Gert-Åke Hanssonb,c, Helen Cristina Nogueiraa, Henrik Enquistb,c,
Andreas Holtermannd, Ana Beatriz Oliveiraa,∗
a
Laboratory of Clinical and Occupational Kinesiology (LACO), Department of Physical Therapy, Federal University of São Carlos, Rodovia Washington Luis, KM 235, CP
676, 13565-905, São Carlos, SP, Brazil
b
Occupational and Environmental Medicine, University and Regional Laboratories Region Scania, SUS, Lund, 221 85 Lund, Sweden
c
Division of Occupational and Environmental Medicine, Lund University, SUS, Lund, 221 85 Lund, Sweden
d
National Research Centre for the Working Environment, Lersø Parkallé 105, DK-2100 Copenhagen, Denmark

A R T I C LE I N FO A B S T R A C T

Keywords: Automated processes in industry have been implemented in order to get higher production rates, but our
Assembly work knowledge about their effects on physical exposure of the workers is still limited. Muscular load, postures, and
Industrial automation movements of the head, neck/shoulders, upper arms, and wrists were recorded in 19 operators from two dif-
Musculoskeletal disorders ferent car clutch disc production lines (manual and semi-automated). Higher angular velocities of the head,
Physical exposure
upper back, and upper arms were found on the manual production line (on average 20% higher than the semi-
Repetitive work
automated). Upper trapezius, and forearm extensor rest (% time), as well as hand kept still (% time) were also
higher on the manual production line. No difference was found regarding posture, muscular activities, and
repetitiveness. The manual line had more rest, but more vigorous movements considering angular velocities. The
semi-automated line, therefore, implied a higher production rate with lower angular velocities but fewer op-
portunities for rest than the manual line. While different physical exposures were found when comparing these
two production lines with different levels of automation, the health effects derived from industrial automation
ought to be investigated with a larger sample size.
Relevance to industry: The reduction of rest opportunities observed on a semi-automated production line requires
the attention of ergonomists when planning or redesigning tasks in such production lines. Lack of rest is con-
sidered a risk factor for the development of musculoskeletal disorders.

1. Introduction human operators are responsible for conducting the tasks. Semi-auto-
mated systems are those in which material handling is still conducted
Modern production systems have increased production rates by in- by human operators, but there is a higher level of automation regarding
troducing new technology. This has been possible with the combination alignment of the materials and the performance of specific tasks on
of skilled human operators and the introduction of automated pro- them by robots. In automated systems, material handling is also auto-
duction processes, which ensures higher production efficiency. mated and involves less interference by human operators (Duncheon,
However, the introduction of more complex mechanization and auto- 2002). The need for human intervention on those different automation
mation may include new biomechanical and psychological demands levels may influence the physical and/or psychological demands to
(Coury et al., 2000; Garde et al., 2003). which workers are exposed. Previous studies have identified that
According to Frohm (2008), the concept of levels of automation can manual production systems are usually characterized by having more
be described as “the allocation of physical and cognitive tasks between diverse work, more exposure variation, and higher levels of alertness
humans and technology, described as a continuum ranging from totally and psychological activation when compared to semi-automated or
manual to totally automatic”. The automated processes are often de- automated production systems (Balogh et al., 2006; Coury et al., 2000;
scribed as manual, semi-automated and/or automated (Duncheon, Garde et al., 2003; Juul-Kristensen et al., 2002; Stål et al., 2003).
2002; Johansson et al., 2009). Manual systems are those in which Automated production systems are frequently followed by

∗
Corresponding author. Universidade Federal de São Carlos, Departamento de Fisioterapia, Rodovia Washington Luis, KM 235, CP 676, 13565-905, São Carlos, SP, Brazil.
E-mail addresses: [Link]@[Link] (F. Locks), [Link]@[Link] (G.-Å. Hansson), helencnogueira@[Link] (H.C. Nogueira),
[Link]@[Link] (H. Enquist), aho@[Link] (A. Holtermann), biaoliveira@[Link], biaoliveira@[Link] (A.B. Oliveira).

[Link]
Received 31 May 2017; Received in revised form 7 April 2018; Accepted 12 April 2018
0169-8141/ © 2018 Elsevier B.V. All rights reserved.
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

rationalization of the manufacturing process, generating more con- from each production line) were excluded.
strained and repetitive work with less opportunity for short breaks The subjects were interviewed about demographic data and mus-
during work (Mathiassen, 2006; Palmerud et al., 2012; Srinivasan and culoskeletal complaints according to the Brazilian version of the Nordic
Mathiassen, 2012). This is of great concern as those characteristics are Questionnaire (Barros and Alexandre, 2003). The operators were also
assumed to be risk factors for the development of musculoskeletal physically examined regarding neck, shoulders, elbows, wrists, and
disorders (MSDs), especially in the neck, shoulder, hands, and upper hands (Ohlsson et al., 1994) and diagnoses were made by the examiner
back (Barbieri et al., 2012; Buckle and Devereux, 2002; Kilbom, 1994; according to predefined criteria (Nordander et al., 2009). All subjects
Sluiter et al., 2001). were examined by the same examiner, a physiotherapist experienced in
Balogh et al. (2006) pointed out that decreasing the biomechanical occupational health assessments. The study is in accordance with the
workload is one goal of the automation and, consequently, it might Declaration of Helsinki and has also been previously approved by the
reduce the risk of MSDs. However, studies focusing on the introduction local Ethics Committee. All participants signed an informed consent
of new technology or mechanization/automation of production systems form before entering the study.
have conflicting results regarding its effect on the biomechanical ex-
posure measurements (Arvidsson et al., 2012, 2006; Balogh et al., 2006; 2.2. Work tasks
Coury et al., 2000; Juul-Kristensen et al., 2002; Stål et al., 2000). In
general, muscular load decreased with the implementation of new On the days of the measurements, recordings were made for about
technology while muscular rest (% time) increased for wooden board seven hours during normal work, including breaks and a mealtime.
production (Balogh et al., 2006) and meat cutters (Arvidsson et al., Each production line had a certain number of distinct tasks and all
2012) but decreased for milk tethering (Stål et al., 2000). More narrow operators performed all tasks at their production line, taking turns. That
posture range and lower movements velocity of the head, neck, upper meant that, for the manual line, the duration of each work task was
arms, and wrists were observed in air traffic controllers (Arvidsson approximately one hour, and, for the semi-automated line, each task's
et al., 2006) and meat cutters (Arvidsson et al., 2012) after introducing duration was approximately 45 min. In total, the measurements were
new technology. On the other hand, Balogh et al. (2006) observed a conducted in a three-month period and were performed by the same
more varied and less constrained movement pattern in an automated researcher. The order of measurements was random between the two
wooden board production line compared to both manual and semi- production lines and the two work shifts.
automated lines. The literature shows that the effects of automation are The tasks performed in each production line are described in Figs. 1
complex, and the conflicting results found cannot support synthesized and 2. The manual line was organized into five different workstations
evidence on this issue. with one operator for each (Fig. 1). Approximately 140 clutch discs
Investigation of the biomechanical workload from manual and au- were produced per hour. On this production line, the operators had to
tomated process requires direct comparison between the two processes. manually handle the clutch disc, position different component parts on
However, most studies have not been able to do so because: (1) the the disc, manually activate the machinery of the production system, and
automation of a work process often influences both the automated and manually move the disc to the next workstation. The work at the semi-
manual production processes, which hampers direct comparison, and automated line was organized as a conveyor belt with seven work-
(2) the workers need to be experienced performing the automatized and stations plus one additional workstation located apart from the con-
the manual work tasks before it can allow for direct comparison. Thus, veyor belt (Fig. 2). Eight operators worked at the conveyor belt, with
in this study, the biomechanical load investigated is made upon a si- two operators in one of the workstations (Task 6 – Fig. 2). The operators
tuation in which the automation process is investigated in two si- at the conveyor belt manually positioned different component parts on
multaneously active production lines with different levels of automa- the disc that was passing through on the conveyor. Different machinery
tion. allowed the work on the semi-automated line to be performed with no
This study aims to investigate potential differences in biomecha- need for manual activation of the machinery. The separate workstation
nical workload between manual and semi-automated production sys- was operated by three operators manually pre-molding one of the
tems of automotive components. We hypothesized that: (1) the manual clutch's parts. The automated apparatus of this production line, apart
production system comprises higher peak load than the semi-automated from the conveyor belt, were composed of automatic robots responsible
production system; (2) the semi-automated system has more repetitive for processes such as pressing and screwing between workstations. The
work and less opportunity for hand kept still and of muscular rest. production rate for the semi-automated line was 260 clutch discs per
hour. The only identical task in the two production lines was the visual
2. Methods inspection, which was the final task in both lines.
The main difference between the production lines is the need for
2.1. Production lines and subjects manual handling the clutch discs and manually pressing the compo-
nents (tasks 2–3) on the manual line. On the semi-automated line, the
The study was conducted in an automotive parts production facility manual handling and pressing processes were replaced by the conveyor
situated in the state of São Paulo, Brazil. Two production lines, one belt and robots between workstations. Take note that the manu-
manual and one semi-automated, were studied. Both production lines facturing processes performed in tasks 1–4 on the manual line were
produce identical car clutch discs, with an average weight of 1.2 kg and equivalent to those performed in tasks 1–6 on the semi-automated line.
a diameter of 200 mm. The two teams of operators from the two pro- On both production lines, the facility implemented a two-hour ro-
duction lines both work eight-hour shifts. tation in which the operators could change tasks according to a pre-
All 43 operators (all male) currently employed at the company and defined scheduled (i.e.: moving from task 1 to task 2, moving from task
working on the manual or the semi-automated lines were invited to and 2 to task 3, etc.). In this way, the operators, on a regular working day,
participated in an information meeting. All operators worked ex- would perform four different tasks. With the aim of getting re-
clusively on one of the lines, and all of them were experienced presentative measurements in a one-day recording, the task rotation
(minimum 6 months). Forty-one operators volunteered for an interview schedule was changed in order to ensure a minimum of 40-min re-
and physical examination. Twenty-seven volunteered to wear the cording of each task on each production line. It meant that, on the
measurement equipment for recording physical workload during one measurement day, each operator performed all tasks from his produc-
workday. Valid measurements were obtained for 19 right-handed op- tion line. This procedure was previously authorized by the facility since
erators (nine from the manual, and ten from the semi-automated line). this new arrangement would not interfere with the production process.
Due to technical issues, the measurements from eight operators (four Both production lines had mobile operators supporting the overall

42
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

Fig. 1. Schematic illustration of the different tasks performed at each workstation on the manual production line.

production. These mobile operators did not take part in the clutch disc 2.3. Direct technical measurements of physical workload
assembly and their work was therefore not addressed in the direct
measurements. All participants were monitored during both work and The biomechanical exposure of the operators from the two pro-
breaks and information such as task duration (start and stop) was re- duction lines was assessed by quantifying the muscle activity of the
corded by a researcher using a custom-made tablet application. upper trapezius and the wrist extensors bilaterally through surface
electromyography, as well as postures and movements of the head,
upper back, upper arms, and wrists through inclinometers and biaxial
flexible goniometers.

Fig. 2. Schematic illustration of the different tasks performed at each workstation on the semi-automated line.

43
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

Bipolar surface electromyography (EMG) recordings of the upper and peak load, respectively (Jonsson, 1982; Veiersted et al., 1990).
trapezius and forearm extensors muscles were bilaterally collected at The inclinometer data were low pass filtered at 5 Hz, the transducer
1024 Hz. Adhesive disposable Ag/AgCl EMG electrodes (H124SG, coordinates were transformed to the coordinates of the body segments,
Kendall™ ARBO, Germany) with a diameter of 24 mm were used. Before and “flexion”, “elevation”, and “angular velocities” were calculated
fixing the electrodes, the skin of the delimited regions was shaved and (Hansson et al., 2001). The accuracy and reproducibility of the in-
cleaned with 70% alcohol. For the upper trapezius, the sensor was clinometers per se are 1.3° and 0.2°, respectively, and therefore they are
placed 2 cm lateral from the middle point between the seventh cervical applicable for measuring occupational movements and postures
vertebra (C7) and the acromion (Mathiassen et al., 1995). Electrodes (Hansson, 2015; Hansson et al., 2006, 2001). For the head, upper back,
were also placed on the belly of the wrists extensor muscles (extensor and upper arms, the 10th, 50th and 90th percentiles of the flexion
carpi radialis longus and brevis) at a distance corresponding to one- angular distributions, elevation angles, and angular velocity distribu-
third of the length of the forearm from the elbow (Hansson et al., 2009). tion were used as exposure measures.
Three maximal voluntary contractions (MVC) were obtained to The goniometry data were low pass filtered at 5 Hz, the angular
normalize EMG signals and were performed as described by Nordander velocity was calculated as the absolute value of the time derivative of
et al. (2004). The upper trapezius MVCs were recorded while the op- the angular data, and the MPF was calculated (Balogh et al., 2009;
erator stood with one shoulder at a time elevated at 90° in the scapular Hansson et al., 1996). The 10th, 50th and 90th percentiles for wrist
plane. The contraction was held for three seconds with a strap, located flexion/extension were used to describe postures. Additionally, the
proximal to the elbow, connected to a force transducer (Kratos, Model percentage of time with angular velocity less than one degree per
DDK100, Brazil) anchored to the floor. Forearm extensors’ MVCs were second (< 1°/s), the 10th, 50th, and 90th percentiles of the angular
obtained with the operator seated, with one upper arm at a time ver- velocity, and MPF were used to describe the hand kept still, angular
tically oriented and elbow flexed at 90°. Maximal handgrips were per- velocity distribution, and repetitiveness, respectively.
formed with no forearm or hand support, using a Jamar Handgrip
Dynamometer (Model 5030J1, Preston, MI, USA). 2.5. Statistical analyses
Postures and movements of the head, upper back, and upper arms
were measured with inclinometers based on triaxial accelerometers All statistical procedures were performed using SPSS, Version 20.0
(LoggerTeknologi HB, Åkarp, Sweden) at a sampling rate of 20 Hz. The
(SPSS Inc. Chicago, IL, USA). Independent t-tests were performed to
sensors were fixed with double-sided adhesive tape to the forehead, to
evaluate differences in physical exposure between the two production
the right of the spine at the level of the spinous processes of C7/T1 and
lines. For all tests, statistical significance was set at 5% (p < 0.05).
at the distal insertion of the deltoid muscle, bilaterally. Reference po-
sitions were recorded to calibrate the system to the neutral position and
to specify the direction of each movement. All procedures regarding 3. Results
placement and calibration were performed according to Hansson et al.
(2006, 2001). Musculoskeletal complaints in the neck/shoulders region (12
Wrist posture and movements were bilaterally recorded using months: 56%; 7 days: 17%), as well as in elbow/hands (12 months:
biaxial flexible goniometers (XG65 Biometrics Ltd, Newport, UK). Both 41%; 7 days: 22%) were reported by a high fraction of the operators
flexion/extension and radial/ulnar deviation angles were recorded at (Table 1). Demographic data and musculoskeletal complaints categor-
128 Hz. The goniometers were affixed with double-sided tape on the ized by production line is also shown in Table 1. No differences were
dorsal side of each wrist. The fixed sensor was located on the midline
between the ulna and radius and the telescopic sensor over the third Table 1
metacarpal bone. In order to define 0° of flexion/extension and radial/ Demographic characteristics of the 41 male operators, based on data from in-
terview and physical examination.
ulnar deviation, a recording was made with the wrist held in an ana-
tomically neutral position, with the operators sitting down and resting Variables Total Manual Semi- p
their lower arms on a table with their palms facing down (Greene and (n = 41) (n = 16) automated
(n = 25)
Heckman, 1994; Heilskov-Hansen et al., 2014). Goniometer and EMG
data were collected using a Mobi8 data logger (TMS International B.V., Age (years); Mean (SD)* 30.7 (6.9) 31.4 (8.8) 30.2 (5.5) 0.59
Enschede, Netherlands). Height (cm); Mean (SD)* 174.2 (5.6) 174.5 (4.6) 174.0 (6.3) 0.76
Weight (kg); Mean (SD)* 77.9 (12.1) 82.9 (13.0) 74.7 (10.4) 0.03
BMI (kg/m2); Mean (SD)* 25.6 (3.3) 27.1 (3.4) 24.6 (9.1) 0.02
2.4. Data processing
Time in Company (Months); 84 (6–180) 84 (6–180) 84 (19–156) 0.72
Median (Min – Max)*
The EMG data were high pass filtered at 10 Hz and the goniometer Time in Actual Job (Months); 48 (6–180) 34 (6–180) 54 (6–156) 0.22
data were resampled to 20 Hz before the data were converted from the Median (Min – Max)*
Mobi8 internal data format to the data format used by EMINGO (the Smokers; n (%)# 2 (5) 2 (12) 0 (0) 0.14
Workers enrolled in leisure 20 (49) 9 (56) 11 (44) 0.32
software for analyzing EMG, INclinometry, and GOniometry), devel- time physical activities
oped by Occupational and Environmental Medicine in Lund, Sweden. more than 3 hours/week; n
The EMG data were digitally band pass (30–400 Hz) and notch fil- (%)#
tered (50 Hz and all harmonics); the reason for choosing 30 Hz as the Complaints; n (%)#
Neck/Shoulders
high-pass frequency is presented in Hansson et al. (2000). The root
12-months prevalence 23 (56) 8 (50) 15 (60) 0.37
mean square value (RMS), mean power frequency (MPF) and an artifact 7-days prevalence 7 (17) 2 (12) 5 (20) 0.43
index (ART) were then calculated for epochs of 1/8 of a second; Elbow/Hands
moreover, the power spectrum was calculated for each minute. These 12-months prevalence 17 (41) 6 (37) 11 (44) 0.62
data were used for inspecting the quality and, when needed, excluding 7-days prevalence 9 (22) 2 (12) 7 (28) 0.31

artifacts. After subtraction of the noise level, the RMS values were
SD = Standard deviation; BMI = Body Mass Index; Min = minimum;
normalized to maximum voluntary effort (MVE), i.e. the maximum RMS Max = maximum; n = number of operators.
activity recorded during the MVCs (Hansson, 2011; Hansson et al., *Comparisons between Manual and semi-automated lines made by independent
1997; Nordander et al., 2000). The fraction of time with no EMG-ac- samples t-test.
tivity (defined as RMS < 0.5% MVE), and the 10th, 50th and 90th #
Comparisons between Manual and semi-automated lines made by chi-squared
percentiles were used to represent muscular rest, and static, median, test.

44
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

Table 2 Table 3
Head, upper back, and upper arms postures and movements (mean, with Wrists postures and movements, and forearm muscular load (Mean, with
standard deviation within parenthesis) during the work on the different pro- standard deviation within parenthesis) during work on the different production
duction lines. lines.
Distribution Manual Line Semi-automated p Distribution Manual Line Semi- p
(Percentile) Line (Percentile) automated
Line
Postures and Movements
Head - Flexion Posture and Movements
Posture (°) 10th −0.5 (11.7) −2.2 (9.3) 0.75 Right Wrist - Flexion
50th 24.8 (10.2) 23.6 (7.9) 0.80 Posture (°) 10th −44.4 (3.7) −42.6 (7.6) 0.55
90th 43.3 (9.6) 38.1 (7.1) 0.22 50th −18.2 (3.3) −20.3 (6.9) 0.44
90th 2.9 (3.4) 0.8 (7.9) 0.45
Angular 10th 2.2 (0.6) 2.0 (0.4) 0.48 Movements
velocity (°/s) 50th 17.5 (3.8) 13.8 (2.0) 0.02 Hand kept still (% – 3.8 (2.4) 1.4 (0.7) 0.02
90th 68.9 (12.8) 54.0 (7.0) 0.01 time)
Angular velocity 10th 1.2 (0.6) 1.9 (0.5) 0.03
Upper Back - Flexion (°/s) 50th 20.6 (3.9) 22.5 (3.1) 0.28
Posture (°) 10th −2.0 (7.2) −0.9 (3.6) 0.68 90th 96.5 (8.7) 99.1 (8.6) 0.53
50th 6.8 (8.4) 6.8 (4.7) 0.98 Repetitiveness – 0.32 (0.04) 0.34 (0.02) 0.16
90th 18.3 (8.3) 16.4 (6.5) 0.59 (Hz)

Angular 10th 1.5 (0.4) 1.5 (0.3) 0.85 Left Wrist - Flexion
velocity (°/s) 50th 13.4 (2.6) 10.6 (2.0) 0.01 Posture (°) 10th −48.8 (7.9) −46.0 (6.6) 0.43
90th 52.0 (7.1) 39.0 (6.9) 0.01 50th −22.9 (6.1) −21.2 (5.0) 0.54
90th −2.8 (7.0) 0.6 (5.5) 0.25
Right Upper Arm - Elevation Movements
Posture (°) 10th 10.3 (3.7) 10.9 (2.4) 0.80 Hand kept still (% – 4.2 (2.5) 1.8 (0.7) 0.03
50th 22.1 (5.6) 25.3 (5.4) 0.41 time)
90th 42.0 (7.7) 51.7 (11.3) 0.14 Angular velocity 10th 0.9 (0.5) 1.3 (0.4) 0.05
(°/s) 50th 16.1 (4.0) 18.3 (3.5) 0.23
Angular 10th 5.9 (2.7) 7.3 (0.5) 0.33 90th 85.0 (10.8) 90.7 (11.5) 0.29
velocity (°/s) 50th 40.8 (4.2) 34.2 (1.7) 0.01 Repetitiveness – 0.32 (0.04) 0.35 (0.03) 0.16
90th 128.2 (10.8) 104.2 (11.6) 0.01 (Hz)

Left Upper Arm - Elevation Muscular Load

Posture (°) 10th 10.7 (4.1) 11.4 (4.7) 0.73 Right Forearm Extensors
50th 23.0 (4.8) 23.6 (8.4) 0.85 Rest (% time) – 11.7 (7.8) 3.2 (1.8) 0.02
90th 43.1 (6.5) 50.3 (10.5) 0.10 Activity (% MVE) 10th 0.7 (0.6) 2.1 (1.1) 0.01
50th 7.1 (5.7) 12.6 (5.2) 0.06
Angular 10th 6.1 (2.5) 7.1 (1.1) 0.30 90th 18.5 (13.5) 29.5 (10.9) 0.08
velocity (°/s) 50th 40.9 (8.7) 33.0 (5.4) 0.03
90th 127.4 (26.5) 102.3 (23.4) 0.04 Left Forearm Extensors
Rest (% time) – 7.1 (4.4) 3.5 (2.3) 0.03
Muscular Load Activity (% MVE) 10th 1.1 (1.0) 1.8 (1.1) 0.20
Right Trapezius 50th 11.0 (5.7) 12.8 (6.7) 0.56
Rest (% time) – 7.2 (7.4) 2.6 (2.2) 0.08 90th 31.0 (14.0) 32.4 (15.5) 0.84
Activity (% 10th 1.8 (1.5) 2.1 (1.0) 0.45
MVE) 50th 5.3 (3.1) 5.7 (1.4) 0.72 Statistically significant differences are marked in bold.
90th 11.5 (3.8) 10.7 (2.0) 0.58

differences in wrist postures between the lines. Repetitiveness showed

Left Trapezius
Rest (% time) – 13.6 (11.2) 5.8 (6.6) 0.08 no statistically significant difference between the two production lines.
Activity (% 10th 1.0 (1.4) 2.0 (2.0) 0.25 Both the right and left forearm extensors showed significantly more rest
MVE) 50th 4.5 (4.0) 5.1 (2.8) 0.75 for the manual line and the right forearm extensors also showed lower
90th 10.8 (5.0) 9.8 (3.8) 0.63
load for the 10th percentile as well as a tendency for lower loads at the
Statistically significant differences are marked in bold.
50th and 90th percentiles. No differences were found regarding radial/
ulnar deviation (data not shown). Furthermore, no difference was found
found between the production lines regarding musculoskeletal com- for the force measurements conducted during the MVC tests
plaints. Only two diagnoses were identified on the physical examina- (p > 0.05). Direct comparisons between tasks within the same pro-
tion: tension neck syndrome – one operator from the semi-automated duction line are shown in Appendix.
line; acromioclavicular syndrome – one operator from each production Analyses investigating the differences between tasks within the
line. same production line are included as Appendices.
Regarding the head, upper back, and upper arms, the manual line
had, for both the 50th and 90th percentiles, the highest velocities of the
4. Discussion
two lines (Table 2). However, the postures showed no statistically sig-
nificant differences between the two production lines. There was, for
The manual line showed higher median velocities for the head,
both the right and left upper trapezius, a tendency towards more
upper back, and upper arms, as well as a tendency towards more
muscular rest at the manual line.
muscular rest for the trapezius muscles compared to the semi-auto-
At the manual line, both hands were kept still for a longer time and
mated line. At the manual line, the hands were kept still for a longer
the 10th percentiles of angular velocity were lower (Table 3). Contrary
time, and the forearm extensors showed significantly more rest than at
to the upper part of the body, there were no noticeable differences
the semi-automated line. There were no noticeable differences between
between the lines in velocity for the 50th and 90th percentiles. As for
the lines regarding median wrist velocity and repetitiveness, nor pos-
the postures of the upper part of the body, there were no significant
tures for the upper part of the body and the wrists.

45
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

4.1. Physical workload 2003; Juul-Kristensen et al., 2002; Stål et al., 2003). However, these
differences become bigger the higher the difference is in the automation
Muscular activity of the trapezius showed no difference between the process. Another factor to consider in this analysis is how the work is
two production lines, which did not support our first hypothesis. On the organized in both production lines. Rationalization of the manu-
other hand, the manual line showed higher angular velocities for head, facturing process is often seen in more automated systems, which can
upper back, and upper arms considering both the 50th and 90th per- lead to more constrained and repetitive work with less opportunity for
centiles. On average, the manual line had 20% higher angular velocity, breaks (Mathiassen, 2006; Palmerud et al., 2012; Srinivasan and
considering all upper body parts, which might be related to more vig- Mathiassen, 2012). One intervention often mentioned as a way to en-
orous movements performed on this production line. Those findings hance variation is job/task rotation (Mathiassen, 2006; Srinivasan and
show that the automation process leads to work that is both less vig- Mathiassen, 2012). In this study, both production lines work similarly
orous and less diverse. in that they have task rotations every two hours. On the other hand, the
We believe that the higher velocities found for the manual line can tasks performed on each production line varied significantly; i.e., on the
be explained by the need to perform more diverse manual work to manual line, there was more varied work between the tasks performed
produce the final clutch disc. While the operators on the manual line while on the semi-automated, even though there are more work-
had to manually handle the clutch discs, position the component parts, stations, the physical load to which the operator was exposed was not
and manually activate the machinery of the production system, the very different between them. In this way, job/task rotation is not the
operators on the semi-automated line only performed the manual po- only important key to create variation. There is also a need to identify
sitioning of the component parts. This is in accordance with Balogh tasks with different workloads to implement a better task rotation
et al. (2006) and Arvidsson et al. (2012) who also found higher angular scheme that allows the worker to experience different biomechanical
velocities for head and upper arms on the most manual system when exposures during the workday.
comparing different automated processes in parquet flooring and meat Some of the findings are comparable to previously observed mea-
cut industries, respectively. However, in both the study of Balogh et al. surements performed on male workers. Fig. 3 is a graphical re-
(2006) and Arvidsson et al. (2012) a higher muscular activity of the presentation of the mean values for both upper arms and wrist variables
trapezius muscles was also found on the most manual system, which in comparison with previously published studies with the same methods
was not observed in our study. The contrasting findings of these pre- (Hansson et al., 2010, 2009).
vious studies and ours can be explained by the work performed since Considering muscular rest of the right trapezius (Fig. 3a), even
producing the clutches does not require static contractions of the though not being statistically different on the two investigated pro-
shoulder in order to perform the tasks, in contrast to parquet flooring duction lines, the lower values observed for the semi-automated line are
(Balogh et al., 2006) and meat cutting (Arvidsson et al., 2012) activ- closely plotted with more demanding work, such as food processing and
ities. assembly work. For the median upper arm velocity (50th percentile –
As first hypothesized, the semi-automated line provided less op- Fig. 3b) the manual line is plotted closely to those previously mentioned
portunities to rest compared to the manual line. Rest was found to be work situations, while the semi-automated work shows a lower upper
consistently higher on the manual line, both for muscular rest and arm velocity. This contrast indicates that the manual line had more
wrists movements. The fraction of time of muscular rest of the trapezius vigorous work (higher velocities) but with more rest opportunities than
(even though not statistically different between the two lines), mus- the semi-automated line.
cular rest of the forearm extensors, and time the hand was kept still For the time the right hand was kept still (Fig. 3c), the lower values
were twice as high in the manual line than in the semi-automated. The observed for the semi-automated line are also plotted closely to in-
muscular rest and the time the hand is kept still are presented as an dustrial tasks and activities demanding repetitive wrist movements (e.g.
indication of risks associated with a lack of recovery, according to parcel sorting, letter sorting). Those are work situation in which low
Nordander et al. (2000) and Arvidsson et al. (2012), respectively. One rest opportunities and highly repetitive manual work is observed.
thing to take into consideration is the different work pace between the Considering the 50th percentile of the right wrist velocity (Fig. 3d),
production lines. On the manual line, the work was self-paced which both production lines had similar results and were found in between the
may have given the operators the opportunity to have more unplanned other work activities shown in comparison.
breaks during the work day. Differently, the semi-automated line had
its pace imposed by the machine, which does not allow them to have 4.2. Musculoskeletal disorders
the opportunity to have unplanned breaks during the work day. An-
other thing to be considered when interpreting these findings is a dif- A high fraction of the operators reported complaints in the neck/
ference in MVE activity for the right forearm extensors between the two shoulders region (12 months: 53%; 7 days: 17%), as well as in elbow/
production lines. The right forearm extensors’ MVE was higher for the hands (12 months: 44%; 7 days: 19%). Nordander et al. (2009) found
operators at the manual line (786 μV) than for the operators at the semi- slightly higher results for the neck/shoulders (12 months: 62%; 7 days:
automated line (439 μV). However, this difference was statistically non- 32%) and elbows/hands (12 months: 56%; 7 days: 27%) regions in a
significant (p = 0.14). Our interpretation is that, although the differ- sample of 622 male industrial workers. The identification of only two
ences between the lines may be somewhat overestimated, our conclu- diagnoses (tension neck syndrome and acromioclavicular syndrome)
sions are not invalidated. This interpretation can be supported by the does not reflect the high prevalence of reported musculoskeletal com-
methodological care during data collection and by no major issues plaints on both neck/shoulders and elbow/hands. This can be attrib-
identified when scrutinizing the relation between the percentile values uted to rigorous diagnostic criteria and the lack of diagnostic entities in
for both hands and both lines. hands (Ohlsson et al., 1994; Sluiter et al., 2001).
Even though differences were found between the two production
lines, when comparing the throughput of both lines per operator per 4.3. Study strengths and limitations
day, the difference between them is small. One thing to consider in this
analysis is the fact that the production system seen in the semi-auto- One strength of this study is the use of technical measurements,
mated line involved a low level of automation. The literature points out since they provide objective and generic metrics of exposure regarding
that more diverse work, more exposure variation, and higher levels of postures, movements and muscular load. This also enables direct
alertness and psychological activation are usually related to manual comparisons of the exposure with other types of jobs and tasks, and
production systems when compared to semi-automated or automated estimating the risk for MSDs, based on studies using the same metrics.
production systems (Balogh et al., 2006; Coury et al., 2000; Garde et al., Another strength of this study is the comparison of two simultaneously

46
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

Fig. 3. Mean exposure values (and 95% confidence intervals) for the right side in groups of male workers performing different types of work. Data from Fig. 3a
(Upper trapezius muscular rest [% time]) and 3b (Upper arm velocity [50th percentile; °/s]) were extracted from Hansson et al. (2010). Data from Fig. 3c (Hand kept
still [% time]) and 3d (Wrist velocity [50th percentile; °/s]) were extracted from Hansson et al. (2009).

active production lines. peak loads or more repetitive work. On the other hand, it provided less
The parameters used in this study to address physical exposure are opportunity for both muscular rest and the hands to be kept still. Even
traditional exposure measures that use the cumulative exposure over though differences were found between the two levels of automation,
the work day to express possible physical workloads. However, they do its health effects, such as musculoskeletal disorders or sickness absence,
not reflect the time pattern of these exposures. In this sense, further still remain to be addressed with a larger epidemiological study.
studies should consider reporting the variation of the biomechanical
exposure, addressed by Mathiassen (2006) as the change in exposure
Funding
across time, and identify possible influences of automation processes on
physical exposure variation.
This work was supported by CAPES Foundation, Ministry of
Education, Brazil (Proc.: 006555/2015-02).
5. Conclusions

The results of this study showed that the semi-automated line with a Acknowledgements
higher production rate provided lower angular velocities of the body,
but with lower rest opportunities compared to the manual line. The Skilled quality control and analysis of the exposure measurements
semi-automated line, as first hypothesized, did not comprise higher was made by Ms. Lothy Granqvist.

47
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

Appendices

Appendix 1
Head, upper Back, and upper arms postures and movements (Mean, with standard deviation within parenthesis) during the different tasks performed
on the Manual Line.

Distribution (Percentile) Task 1 Task 2 Task 3 Task 4 Task 5

Postures and Movements

Head - Flexion
Posture (°) 10th −2.2 (8.8) 7.7 (15.3) 5.4 (16.4) −1.9 (13.8) 1.1 (12.9)
50th 16.8 (9.4) 2,3 33.9 (11.2) 1,4
33.3 (11.9) 1,4
16.8 (14.1) 2,3 23.0 (12.2)
90th 34.5 (8.6) 45.2 (9.8) 45.2 (9.7) 37.5 (11.7) 44.3 (11.5)

Angular velocity (°/s) 10th 2.0 (0.7) 1 2.6 (0.7) 2.4 (0.7) 3.4 (1.5) 1,5 1.8 (0.4) 1
50th 15.8 (5.4) 1 18.1 (4.8) 1 17.0 (3.5) 1 24.8 (6.7) 1,2,3,5 15.0 (3.1) 1
90th 63.8 (14.7) 1 67.8 (18.6) 65.8 (12.5) 83.3 (16.5) 1,5 61.9 (11.8) 1

Upper Back - Flexion

Posture (°) 10th −2.0 (6.7) 1.6 (8.7) 0.6 (8.8) −2.6 (7.0) −3.0 (7.9)
50th 4.5 (7.1) 9.4 (9.3) 9.2 (10.2) 6.4 (7.3) 4.0 (9.0)
90th 13.6 (7.0) 18.1 (9.9) 17.7 (9.9) 19.0 (8.2) 18.7 (9.2)

Angular velocity (°/s) 10th 1.4 (0.5) 4 2.0 (0.5) 1.9 (0.5) 2.7 (1.3) 1 1.2 (0.2)
50th 11.6 (4.2) 4 14.2 (3.3) 4 13.4 (2.8) 4
20.5 (5.7) 1,2,3,5 10.6 (2.2) 4

90th 46.6 (11.1) 4 49.8 (10.7) 4 49.0 (7.3) 4

66.5 (11.0) 1,2,3,5 44.4 (6.9) 4

Right Upper Arm - Elevation

Posture (°) 10th 10.0 (3.6) 13.6 (5.2) 12.6 (4.5) 7.8 (2.8) 11.3 (3.8)
50th 19.2 (3.5) 24.3 (6.8) 26.2 (8.8) 19.3 (5.5) 23.5 (7.0)
90th 35.1 (5.9) 42.2 (7.2) 46.4 (10.1) 40.5 (8.8) 42.9 (11.6)

Angular velocity (°/s) 10th 5.2 (3.2) 9.5 (3.6) 8.8 (3.5) 11.7 (7.7) 5 4.1 (1.3) 4
50th 34.6 (16.4) 4 40.4 (7.1) 4 44.0 (8.7) 60.8 (14.9) 1,2,5 33.5 (4.2) 4
90th 113.8 (33.5) 4 115.8 (16.7) 4
132.2 (26.7) 159.9 (28.9) 1,2,5 111.8 (9.4) 4

Left Upper Arm - Elevation

Posture (°) 10th 9.8 (4.0) 14.3 (5.1) 12.2 (4.5) 8.9 (3.5) 12.0 (4.9)
50th 20.8 (5.5) 27.0 (6.1) 23.3 (4.8) 20.7 (3.8) 23.8 (4.3)
90th 38.6 (6.6) 47.6 (7.2) 41.5 (5.5) 42.9 (9.0) 42.4 (10.5)

Angular velocity (°/s) 10th 5.3 (2.5) 2,4 10.5 (3.9) 1,5 8.0 (2.6) 11.8 (7.4) 1,5 3.8 (1.1) 2,4
50th 37.1 (13.7) 4 42.3 (11.9) 37.5 (7.3) 4 57.8 (15.1) 1,3,5 32.5 (10.0) 4
90th 119.1 (32.0) 134.1 (32.1) 111.2 (22.2) 155.4 (36.8) 110.9 (29.8)

Muscular Load
Right Trapezius
Rest (% time) – 12.9 (15.3) 2.5 (3.4) 3.4 (2.6) 7.1 (8.6) 7.2 (10.9)
Activity (%MVE) 10th 1.2 (1.1) 3.1 (2.1) 2.9 (2.1) 1.3 (1.0) 2.5 (2.2)
50th 3.5 (2.3) 3 7.2 (2.8) 8.0 (3.7) 1 4.1 (2.5) 5.3 (3.7)
90th 7.8 (3.6) 3 12.2 (3.6) 13.7 (4.6) 1 10.1 (3.1) 10.6 (4.8)

Left Trapezius
Rest (% time) – 18.5 (13.5) 3.7 (3.8) 6.6 (7.5) 17.0 (18.4) 22.0 (25.8)
Activity (% MVE) 10th 0.5 (0.6) 2.9 (3.2) 2.2 (2.3) 0.8 (0.9) 1.0 (1.1)
50th 2.9 (2.7) 7.1 (4.2) 6.6 (4.2) 4.1 (4.5) 3.9 (3.9)
90th 6.7 (4.1) 12.7 (4.4) 11.3 (5.5) 9.2 (6.9) 9.2 (5.9)

*Superscripted numbers indicate statistically significant difference (P < 0.05) with the task with the same number.

48
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

Appendix 2
Wrists posture, movements and forearm muscular load (Mean, with standard deviation within parenthesis) on the different tasks performed on the
Manual Line

Distribution (Percentile) Task 1 Task 2 Task 3 Task 4 Task 5

Posture and Movements

Right Wrist - Flexion
Posture (°) 10th −45.4 (6.2) −38.8 (7.0) −41.4 (6.2) −46.4 (8.2) −46.0 (7.5)
50th −19.4 (3.1) −13.5 (4.2) −14.5 (4.2) −21.0 (6.3) −23.9 (4.6)
90th −1.0 (2.7) 6.9 (5.9) 7.3 (5.6) −0.8 (9.2) −3.6 (5.7)
Movements
Hand kept still (% time) – 6.7 (6.7) 1.2 (1.5) 1.9 (1.4) 3.2 (3.9) 4.3 (4.0)
Angular velocity (°/s) 10th 0.8 (0.6) 2.4 (1.1) 2.4 (1.4) 2.0 (1.4) 1.0 (0.7)
50th 14.7 (7.1) 2,3 24.8 (5.0) 2 28.0 (6.7) 2,5 23.8 (7.9) 16.5 (3.8) 3
90th 83.4 (16.7) 4 93.2 (10.7) 103.5 (16.7) 111.7 (21.0) 2
90.4 (11.5)
Repetitiveness (Hz) – 0.32 (0.03) 0.35 (0.06) 0.38 (0.07) 0.44 (0.13) 0.35 (0.07)

Left Wrist - Flexion

Posture (°) 10th −43.2 (8.3) −45.8 (11.9) −45.5 (8.0) −50.9 (8.3) −54.6 (7.9)
50th −18.6 (5.4) −21.8 (7.2) −20.8 (7.7) −26.6 (10.7) −28.3 (7.4)
90th −2.9 (7.7) −1.1 (6.8) −0.4 (7.9) −5.9 (11.8) −8.6 (5.7)
Movements
Hand kept still (% time) – 7.4 (7.0) 2 1.4 (1.7) 1 2.3 (1.2) 3.1 (3.5) 5.0 (3.7)
Angular velocity (°/s) 10th 0.6 (0.4) 2 2.0 (0.9) 1 1.2 (0.4) 1.6 (1.0) 0.7 (0.4)
50th 13.0 (5.2) 21.4 (5.7) 5 16.6 (3.5) 19.0 (5.3) 12.4 (3.8) 2
90th 77.9 (18.4) 90.4 (12.4) 83.6 (11.2) 90.8 (13.9) 81.4 (11.8)
Repetitiveness (Hz) –

Muscular Load
Right Forearm Extensors
Rest (% time) – 18.5 (12.1) 5.3 (6.1) 7.3 (8.2) 12.1 (9.5) 15.1 (9.6)
Activity (% MVE) 10th 0.3 (0.3) 2.9 (3.0) 3.3 (3.6) 0.6 (0.6) 0.7 (1.0)
50th 4.6 (3.7) 9.0 (7.5) 9.7 (8.4) 6.0 (4.6) 5.2 (4.5)
90th 15.8 (12.1) 18.7 (14.8) 21.1 (16.7) 18.8 (12.3) 15.5 (8.8)

Left Forearm Extensors

Rest (% time) – 14.2 (9.5) 2,3,4 2.1 (2.0) 1 2.7 (3.7) 1 4.5 (2.2) 1 11.0 (8.8)
Activity (% MVE) 10th 0.5 (0.5) 4.0 (3.7) 3.4 (3.0) 1.7 (1.7) 1.0 (1.2)
50th 7.4 (6.1) 15.3 (8.0) 13.7 (8.0) 10.4 (6.0) 9.0 (5.4)
90th 26.6 (16.1) 33.4 (15.1) 31.0 (15.5) 31.6 (17.6) 34.3 (14.3)

*Superscripted numbers indicate statistically significant difference (P < 0.05) with the task with the same number.

Appendix 3
Head, upper Back, and upper arms postures and movements (Mean, with standard deviation within parenthesis) during the different tasks performed
on the Semi-automated Line

Distribution Task 1 Task 2 Task 3 Task 4 Task 5 Task 6 Task 7 Task 8

(Percentile)

Postures and Movements

Head - Flexion
Posture (°) 10th −0.7 (9.6) −1.5 (10.3) 1.2 (9.6) −3.7 (10.0) −1.3 (9.8) 9.8 (9.5) −0.1 (12.9) 4.4 (8.8)
50th 21.5 (7.5) 21.7 (7.4) 25.3 (7.1) 14.4 (9.7) 6 21.6 (9.8) 31.3 (7.7) 4
24.6 (9.0) 26.4 (8.4)
90th 33.8 (7.3) 33.9 (7.3) 37.0 (7.7) 31.0 (7.2) 34.7 (8.7) 39.5 (8.2) 39.8 (5.7) 45.6 (11.6)

Angular 10th 1.7 (0.3) 2.2 (0.3) 2.3 (0.6) 2.4 (0.5) 2.6 (0.6) 1.9 (0.6) 1.8 (0.4) 1.9 (0.3)
velocity (°/s) 50th 11.5 (1.8) 14.1 (2.0) 14.6 (3.5) 16.6 (2.1) 16.9 (3.4) 11.6 (3.4) 12.1 (2.4) 13.1 (1.7)
4,5
90th 43.3 (7.2) 52.0 (7.9) 50.7 (9.3) 63.7 (9.8) 1
64.7 (12.2) 1,6 45.0 (12.4) 5
48.6 (8.9) 53.0 (10.4)

Upper Back - Flexion

Posture (°) 10th −0.5 (2.9) 0.8 (4.0) −0.1 (3.7) −3.1 (4.8) −1.0 (5.0) 4.6 (6.0) −0.4 (3.7) −0.3 (3.2)
50th 5.2 (3.3) 6.8 (4.6) 7.5 (5.3) 4.0 (6.4) 7.0 (6.3) 11.0 (7.3) 8.5 (5.1) 5.0 (4.0)
90th 12.3 (4.7) 15.4 (6.1) 15.2 (7.2) 12.8 (7.9) 16.1 (8.0) 16.3 (8.4) 20.3 (6.8) 25.9 (19.7)

10th 1.2 (0.2) 1.6 (0.3) 1.8 (0.6) 1.9 (0.7) 2.0 (0.6) 1.4 (0.5) 1.5 (0.3) 1.3 (0.2)
(continued on next page)

49
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

Appendix 3 (continued)

Distribution Task 1 Task 2 Task 3 Task 4 Task 5 Task 6 Task 7 Task 8

(Percentile)

Angular 50th 8.7 (1.7) 4 10.8 (2.0) 11.5 (3.3) 13.6 (3.5) 1 13.5 (3.2) 8.9 (2.8) 9.9 (1.7) 9.9 (1.2)
velocity (°/s) 90th 32.4 (6.6) 4,5 36.9 (6.2) 37.1 (9.0) 50.2 (14.3) 49.2 (10.7) 1,6 32.6 (10.6) 4,5 36.6 (5.3) 39.4 (8.2)
1,6

Right Upper Arm - Elevation

2 2 2 2
Posture (°) 10th 8.9 (4.7) 25.5 (14.1) 12.4 (2.7) 14.6 (4.8) 14.2 (6.0) 13.0 (1.8) 10.5 (3.2) 9.3 (3.1)
1,3,7,8

2 2
50th 25.2 (8.7) 43.1 (16.1) 26.5 (6.8) 31.5 (9.5) 27.8 (8.6) 21.7 (3.5) 20.4 (5.5) 17.2 (5.3) 2
6,7,8

8
90th 51.8 (11.0) 58.7 (11.0) 47.8 (7.3) 54.0 (16.7) 43.6 (10.4) 40.6 (8.2) 42.5 (11.8) 30.8 (4.3) 2

Angular 10th 6.1 (3.2) 7.4 (1.2) 9.1 (2.9) 7.9 (3.5) 8.2 (2.4) 9.6 (1.6) 7.4 (1.8) 6.4 (2.0)
velocity (°/s) 50th 30.7 (7.6) 33.6 (5.4) 35.8 (5.5) 39.1 (8.5) 36.9 (3.1) 34.0 (5.4) 32.3 (5.7) 33.8 (3.5)
90th 95.3 (19.8) 108.7 (25.6) 104.5 (15.2) 119.0 (36.7) 107.8 (13.3) 96.7 (15.2) 96.4 (13.3) 104.4
(15.4)

Left Upper Arm - Elevation

Posture (°) 10th 11.2 (4.8) 18.6 (10.2) 13.0 (4.3) 16.5 (9.1) 15.3 (8.6) 13.5 (6.1) 9.5 (3.2) 9.2 (4.9)
2,4 8 2,4 2,4
50th 21.0 (8.7) 35.8 (13.0) 25.1 (6.7) 34.8 (10.5) 30.8 (13.2) 20.2 (6.6) 19.2 (5.7) 17.5 (8.1)
1,6,7,8 1,6,7,8 2,4,5

8 2,4 2,4
90th 42.1 (13.1) 59.3 (7.7) 46.4 (9.0) 57.8 (12.6) 51.8 (14.0) 37.5 (7.8) 37.5 (12.4) 30.5 (13.3)
6,7,8 6,7,8 2,4,5

Angular 10th 5.8 (0.3) 7.6 (1.8) 9.0 (2.5) 9.4 (5.1) 8.9 (3.0) 7.9 (2.4) 7.3 (1.5) 5.9 (2.3)
velocity (°/s) 50th 27.8 (7.1) 34.2 (7.6) 35.8 (6.7) 41.0 (14.4) 38.5 (8.1) 28.8 (7.5) 31.6 (5.5) 31.2 (6.2)
90th 87.3 (23.9) 105.5 (32.5) 105.3 (29.5) 120.2 (45.3) 114.0 (30.6) 85.9 (23.9) 96.0 (17.3) 98.4 (18.7)

Muscular Load
Right Trapezius
Rest (% time) – 3.5 (4.5) 2.5 (3.4) 1.8 (1.3) 0.8 (1.0) 1.0 (0.8) 0.4 (0.4) 5.2 (5.2) 4.1 (4.4)
Activity (% 10th 1.9 (1.2) 2.5 (1.2) 2.5 (1.0) 3.3 (0.9) 2.7 (1.2) 3.8 (1.7) 7 1.5 (0.9) 6
1.8 (1.1)
MVE) 50th 5.0 (1.7) 6.7 (1.4) 6.2 (1.7) 7.2 (1.7) 6.5 (1.9) 6.9 (2.0) 4.6 (1.4) 4.4 (2.0)
90th 9.2 (2.2) 11.5 (2.2) 10.7 (2.4) 11.9 (2.9) 10.9 (2.7) 10.9 (2.6) 9.3 (1.5) 8.1 (2.6)

Left Trapezius
Rest (% time) – 10.5 (14.2) 6.3 (7.6) 2.6 (2.5) 4.6 (8.4) 2.3 (2.9) 1.2 (2.7) 8.4 (9.1) 8.5 (11.1)
Activity (% 10th 1.8 (2.2) 2.0 (1.9) 2.9 (2.3) 2.5 (1.8) 2.9 (2.2) 3.7 (2.5) 1.8 (2.0) 1.6 (1.7)
MVE) 50th 4.1 (2.9) 5.6 (2.9) 6.0 (2.9) 5.8 (2.9) 6.4 (3.4) 6.1 (3.2) 4.2 (2.8) 3.7 (2.8)
90th 7.9 (3.5) 10.1 (3.9) 10.1 (3.9) 10.6 (4.0) 11.4 (5.0) 9.6 (4.2) 8.1 (3.5) 6.7 (3.8)

*Superscripted numbers indicate statistically significant difference (P < 0.05) with the task with the same number.

Appendix 4
Wrists posture, movements and forearm muscular load (Mean, with standard deviation within parenthesis) on the different tasks performed on the
Semi-automated Line.

Distribution Task 1 Task 2 Task 3 Task 4 Task 5 Task 6 Task 7 Task 8

(Percentile)

Posture and Movements

Right Wrist - Flexion
Posture (°) 10th −42.2 (7.6) −38.5 (8.1) −43.0 (9.2) −39.6 (9.8) −42.2 −36.7 (8.1) −45.0 (11.3) −45.8 (7.7)
(10.5)
50th −21.9 (8.3) −18.4 (9.0) −21.8 (8.7) −19.3 (7.5) −19.0 (7.9) −17.9 (7.7) −19.3 (7.2) −24.9 (9.6)
90th −1.4 (11.4) −0.4 (9.4) 0.8 (12.7) 3.4 (10.3) 2.2 (6.8) 0.8 (9.2) 2.9 (10.1) −7.6 (9.7)
Movements
Hand kept still – 2.2 (1.2) 1.1 (0.7) 1.3 (1.2) 1.5 (1.5) 1.2 (1.1) 0.1 (0.2) 1.1 (0.8) 2.5 (2.9)
(% time)
Angular 10th 1.2 (0.4) 4,5,6 2.1 (0.7) 2.5 (1.2) 3.0 (1.5) 1 2.7 (1.3) 1 3.2 (0.9) 1 1.7 (0.7) 1.5 (0.9)
velocity (°/s) 50th 17.7 (3.8) 4,5 23.0 (4.4) 23.6 (5.4) 29.1 (6.2) 1,7,8 27.3 (6.2) 1 25.4 (5.6) 19.3 (5.5) 4 20.0 (7.1) 4
90th 93.3 (16.0) 101.6 (15.3) 98.1 (9.8) 111.1 (16.0) 112.8 (19.3) 89.4 (16.3) 90.9 (14.6) 95.7 (21.1)
–
(continued on next page)

50
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

Appendix 4 (continued)

Distribution Task 1 Task 2 Task 3 Task 4 Task 5 Task 6 Task 7 Task 8

(Percentile)

Repetitiveness
(Hz)

Left Wrist - Flexion

Posture (°) 10th −46.3 (10.6) −41.9 (7.5) −42.0 (5.9) −43.1 (8.1) −43.0 (7.7) −38.7 (3.9) −50.5 (11.9) −48.7
(14.5)
50th −22.8 (6.5) −19.8 (8.3) −23.3 (6.1) −20.0 (7.0) −19.0 (6.5) −19.8 (4.7) −18.1 (5.3) −29.0
(13.0)
90th −2.7 (7.4) −0.1 (8.8) −4.0 (5.2) 1.9 (7.8) 5.7 (7.5) 8
−1.8 (6.7) 3.9 (5.6) 8
−11.0
(10.1) 5,7
Movements
Hand kept still – 2.6 (1.1) 1.5 (1.0) 1.5 (1.5) 1.8 (2.0) 1.7 (1.8) 0.8 (0.7) 1.7 (0.9) 2.6 (2.3)
(% time)
Angular 10th 1.0 (0.6) 4 1.8 (0.7) 1.8 (0.9) 2.5 (1.7) 1,7 1.8 (1.0) 1.7 (0.7) 1.2 (0.4) 4 1.2 (0.6)
velocity (°/s) 50th 13.4 (5.3) 4 20.1 (6.2) 21.1 (4.1) 25.7 (8.3) 1,7,8
21.0 (6.9) 18.2 (5.4) 16.4 (4.1) 4 15.7 (5.1) 4
90th 81.5 (15.2) 5 87.8 (18.7) 96.4 (15.0) 106.2 (19.8) 6 109.6 (25.2) 80.2 (17.9) 4,5 82.8 (14.3) 5 84.0 (17.8) 5
1,6,7,8

Repetitiveness –
(Hz)

Muscular Load
Right Forearm Extensors
Rest (% time) – 3.8 (3.5) 1.7 (1.7) 1.3 (1.2) 2.5 (2.9) 3.4 (2.2) 0.9 (1.3) 5.4 (3.3) 4.7 (4.5)
Activity (% 10th 1.2 (0.6) 6 3.2 (1.8) 3.9 (1.7) 4.0 (2.5) 2.6 (1.8) 4.8 (2.7) 1,7 1.5 (1.3) 6 2.4 (2.1)
MVE) 50th 9.1 (3.3) 13.1 (6.2) 14.1 (6.0) 15.9 (7.1) 13.4 (5.5) 12.5 (5.4) 10.8 (5.7) 11.4 (6.1)
90th 25.6 (8.2) 28.7 (14.1) 29.1 (12.0) 34.4 (14.8) 29.7 (10.5) 24.5 (8.7) 30.6 (11.4) 31.2 (11.8)

Left Forearm Extensors

Rest (% time) – 4.2 (4.1) 2.0 (2.4) 2.0 (2.5) 2.2 (2.9) 5.0 (6.4) 1.5 (1.5) 5.4 (3.4) 4.3 (3.1)
Activity (% 10th 1.0 (0.5) 3.2 (2.6) 3.4 (2.9) 3.8 (3.2) 2.0 (1.7) 3.3 (2.0) 1.1 (0.7) 2.1 (1.8)
MVE) 50th 8.8 (3.7) 14.7 (10.6) 14.6 (7.2) 15.9 (7.8) 13.4 (7.8) 12.3 (7.0) 11.5 (4.6) 11.4 (6.6)
90th 28.0 (13.6) 32.4 (20.8) 32.7 (16.2) 36.9 (17.8) 33.0 (15.1) 26.1 (13.8) 32.2. 13.3) 33.4 (15.3)

*Superscripted numbers indicate statistically significant difference (P < 0.05) with the task with the same number.

References Health 19, 297–306. [Link]

Greene, W.B., Heckman, J.D., 1994. The Clinical Measurement of Joint Motion, first ed.
American Academy of Orthopaedic Surgeons, Rosemont, IL, USA.
Arvidsson, I., Balogh, I., Hansson, G.-Å., Ohlsson, K., Åkesson, I., Nordander, C., 2012. Hansson, G.-Å., 2015. Letter to the editor. Appl. Ergon. 48, 109–110. [Link]
Rationalization in meat cutting - consequences on physical workload. Appl. Ergon. 10.1016/[Link].2014.11.011.
43, 1026–1032. [Link] Hansson, G.-Å., 2011. Letter to the editor. J. Biomech. 44, 1637–1638. [Link]
Arvidsson, I., Hansson, G.-Å., Mathiassen, S.E., Skerfving, S., 2006. Changes in physical 10.1016/[Link].2010.09.008.
workload with implementation of mouse-based information technology in air traffic Hansson, G.-Å., Arvidsson, I., Ohlsson, K., Nordander, C., Mathiassen, S.E., Skerfving, S.,
control. Int. J. Ind. Ergon. 36, 613–622. [Link] Balogh, I., 2006. Precision of measurements of physical workload during standar-
002. dised manual handling. Part II: inclinometry of head, upper back, neck and upper
Balogh, I., Ohlsson, K., Hansson, G.-Å., Engström, T., Skerfving, S., 2006. Increasing the arms. J. Electromyogr. Kinesiol. 16, 125–136. [Link]
degree of automation in a production system: consequences for the physical work- 2005.06.009.
load. Int. J. Ind. Ergon. 36, 353–365. [Link] Hansson, G.-Å., Asterland, P., Holmer, N.G., Skerfving, S., 2001. Validity and reliability of
009. triaxial accelerometers for inclinometry in posture analysis. Med. Biol. Eng. Comput.
Balogh, I., Ohlsson, K., Nordander, C., Skerfving, S., Hansson, G.-Å., 2009. Precision of 39, 405–413.
measurements of physical workload during standardized manual handling part III: Hansson, G.-Å., Asterland, P., Skerfving, S., 1997. Acquisition and analysis of whole-day
goniometry of the wrists. J. Electromyogr. Kinesiol. 19, 1005–1012. [Link] electromyographic field recordings. In: Hermens, H.J., Hägg, G.M., Freriks, B. (Eds.),
org/10.1016/[Link].2008.07.003. Proceedings of the Second General SENIAM (Surface EMG for Non Invasive
Barbieri, D.F., Nogueira, H.C., Bergamin, L.J., Oliveira, A.B., 2012. Physical and psy- Assessment of Muscles) Workshop, pp. 19–27 Stockholm, Sweden.
chosocial indicators among office workers from public sector with and without Hansson, G.-Å., Balogh, I., Ohlsson, K., Granqvist, L., Nordander, C., Arvidsson, I.,
musculoskeletal symptoms. Work 41 (Suppl. 1), 2461–2466. [Link] Åkesson, I., Unge, J., Rittner, R., Strömberg, U., Skerfving, S., 2010. Physical work-
3233/WOR-2012-0481-2461. load in various types of work: Part II. Neck, shoulder and upper arm. Int. J. Ind.
Barros, E.N.C., Alexandre, N.M.C., 2003. Cross-cultural adaptation of the Nordic mus- Ergon. 40, 267–281. [Link]
culoskeletal questionnaire. Int. Nurs. Rev. 50, 101–108. Hansson, G.-Å., Balogh, I., Ohlsson, K., Granqvist, L., Nordander, C., Arvidsson, I.,
Buckle, P.W., Devereux, J.J., 2002. The nature of work-related neck and upper limb Åkesson, I., Unge, J., Rittner, R., Strömberg, U., Skerfving, S., 2009. Physical work-
musculoskeletal disorders. Appl. Ergon. 33, 207–217. load in various types of work: Part I. Wrist and forearm. Int. J. Ind. Ergon. 39,
Coury, H.J.C.G., Léo, J.A., Kumar, S., 2000. Effects of progressive levels of industrial 221–233. [Link]
automation on force and repetitive movements of the wrist. Int. J. Ind. Ergon. 25, Hansson, G.-Å., Balogh, I., Ohlsson, K., Rylander, L., Skerfving, S., 1996. Goniometer
587–595. [Link] measurement and computer analysis of wrist angles and movements applied to oc-
Duncheon, C., 2002. Product miniaturization requires automation – but with a strategy. cupational repetitive work. J. Electromyogr. Kinesiol. 6, 23–35. [Link]
Assemb. Autom. 22, 16–20. [Link] 1016/1050-6411(95)00017-8.
Frohm, J., 2008. Levels of Automation in Production Systems. Thesis Degree Dr. Philos. Hansson, G.-Å., Nordander, C., Asterland, P., Ohlsson, K., Strömberg, U., Skerfving, S.,
Chlamers University of Technology[Link] Rempel, D., 2000. Sensitivity of trapezius electromyography to differences between
Garde, A.H., Hansen, Å.M., Persson, R., Ohlsson, K., Ørbæk, P., 2003. The influence of work tasks - influence of gap definition and normalisation methods. J. Electromyogr.
production systems on physiological responses measured in urine and saliva. Stress Kinesiol. 10, 103–115.

51
F. Locks et al. International Journal of Industrial Ergonomics 67 (2018) 41–52

Heilskov-Hansen, T., Wulff Svendsen, S., Frølund Thomsen, J., Mikkelsen, S., Hansson, G.- physical exposure: the impact of work tasks and individual related factors.
Å., 2014. Sex differences in task distribution and task exposures among Danish house Ergonomics 43, 1904–1919. [Link]
painters: an observational study combining questionnaire data with biomechanical Nordander, C., Ohlsson, K., Akesson, I., Arvidsson, I., Balogh, I., Hansson, G.-A.,
measurements. PLoS One 9. [Link] Strömberg, U., Rittner, R., Skerfving, S., 2009. Risk of musculoskeletal disorders
Johansson, B., Fasth, Å., Stahre, J., Heilala, J., Leong, S., Lee, Y.T., Riddick, F., 2009. among females and males in repetitive/constrained work. Ergonomics 52,
Enabling flexible manufacturing systems by using level of automation as design 1226–1239. [Link]
parameter. In: Rossetti, M.D., Hill, R.R., Johansson, B., Dunkin, A., Ingalls, R.G. Ohlsson, K., Attewell, R.G., Johnsson, B., Ahlm, A., Skerfving, S., 1994. An assessment of
(Eds.), 2009 Winter Simulation Conference, pp. 2176–2184 Austin, TX, USA. neck and upper extremity disorders by questionnaire and clinical examination.
Jonsson, B., 1982. Measurement and evaluation of local muscular strain in the shoulder Ergonomics 37, 891–897. [Link]
during constrained work. J. Hum. Ergol. (Tokyo) 11, 73–88. Palmerud, G., Forsman, M., Neumann, W.P., Winkel, J., 2012. Mechanical exposure im-
Juul-Kristensen, B., Fallentin, N., Hansson, G.-Å., Madeleine, P., Andersen, J., Ekdahl, C., plications of rationalization: a comparison of two flow strategies in a Swedish man-
2002. Physical workload during manual and mechanical deboning of poultry. Int. J. ufacturing plant. Appl. Ergon. 43, 1110–1121. [Link]
Ind. Ergon. 29, 107–115. [Link] 2012.04.001.
Kilbom, A., 1994. Assessment of physical exposure in relation to work-related muscu- Sluiter, J.K., Rest, K.M., Frings-Dresen, M.H., 2001. Criteria document for evaluating the
loskeletal disorders–what information can be obtained from systematic observations? work-relatedness of upper-extremity musculoskeletal disorders. Scand. J. Work.
Scand. J. Work. Environ. Health 20 Spec No, 30–45. Environ. Health 27 (Suppl. 1), 1–102. [Link]
Mathiassen, S.E., 2006. Diversity and variation in biomechanical exposure: what is it, and Srinivasan, D., Mathiassen, S.E., 2012. Motor variability in occupational health and
why would we like to know? Appl. Ergon. 37, 419–427. [Link] performance. Clin. Biomech. (Bristol, Avon) 27, 979–993. [Link]
apergo.2006.04.006. 1016/[Link].2012.08.007.
Mathiassen, S.E., Winkel, J., Hägg, G.M., 1995. Normalization of surface EMG amplitude Stål, M., Hansson, G.-Å., Moritz, U., 2000. Upper extremity muscular load during machine
from the upper trapezius muscle in ergonomic studies - a review. J. Electromyogr. milking. Int. J. Ind. Ergon. 26, 9–17. [Link]
Kinesiol. 5, 197–226. 00059-1.
Nordander, C., Balogh, I., Mathiassen, S.E., Ohlsson, K., Unge, J., Skerfving, S., Hansson, Stål, M., Pinzke, S., Hansson, G.-Å., Kolstrup, C., 2003. Highly repetitive work operations
G.-Å., 2004. Precision of measurements of physical workload during standardised in a modern milking system. A case study of wrist positions and movements in a
manual handling. Part I: surface electromyography of m. trapezius, m. infraspinatus rotary system. Ann. Agric. Environ. Med. 10, 67–72.
and the forearm extensors. J. Electromyogr. Kinesiol. 14, 443–454. [Link] Veiersted, K.B., Westgaard, R.H., Andersen, P., 1990. Pattern of muscle activity during
10.1016/[Link].2003.12.003. stereotyped work and its relation to muscle pain. Int. Arch. Occup. Environ. Health
Nordander, C., Hansson, G.-Å., Rylander, L., Asterland, P., Byström, J.U., Ohlsson, K., 61, 31–41. [Link]
Balogh, I., Skerfving, S., 2000. Muscular rest and gap frequency as EMG measures of

52