Table 2 Workers’ population studies

(1) Humadi et al. 2020 [67]

10 Healthy workers

Cross-sectional case series

(1) Investigating the validity and repeatability of an IMU system for in-field Rapid Upper Limb Assessment (RULA) score assessment in manual handling tasks using 3D Cardan angles and 2D projection angles with respect to values of a motion-capture camera system (reference)

IMU: shoulders, forearms, and wrists

Motion markers: Shoulders (scapula and humerus), and forearms

(1) 6 IMUs (MTw, Xsens

Technologies, the Netherlands, sampling frequency = 60Hz)

(2) Arduino microcontroller

(3) A motion-capture camera system (VICON, Oxford Metrics Group, UK) with eight cameras (sampling frequency = 100Hz)

(1) Root-mean-square error (RMSE) of obtained joint angles (°)

(2) RULA score (1 to 7)

[Using 3 axes upper arm, lower arm and elbow joint angles (°)]

Self-developed software/ Xsens-MVN Analyze

×/✓

(2) Humadi et al. 2021 [68]

11 Healthy novice workers

Cross-sectional case series

(1) Investigating the accuracy and reliability of (a) IMU-based wearable technology and (b) a marker-less optical technology (using Kinect V2) against VICON motion-capture system (the gold standard) based on RULA risk assessment tool during different manual material handling tasks

IMUs: Upper arms, forearms, and hands

VICON markers: shoulders, elbows, and wrists

(1) 6 IMUs (MTw, Xsens Technologies, the Netherlands, sampling frequency = 60 Hz) fixed with plates and attached with using double-sided medical tape and extra layers of surgical tape

(2) Kinect V2 (Microsoft Corporation, USA)

(3) Motion-capture camera system of VICON (Oxford Metrics Group, UK) with eight cameras (sampling frequency = 100 Hz)

(1) Median of joint angles in 3 axes (yaw, pitch, roll; °)

(2) Root-mean-square error (RMSE) of obtained joint

angles in IMU and Kinect methods with respect to VICON system as the gold standard (°)

Self-developed/ Xsens-MVN Analyze

×/✓

(3) Lee et al. 2020 [75]

4 Healthy observers/ subjects

Cross-sectional case series

(1) Testing the interrater reliability of the IMU-based posture-matching method, for providing applications for the IMU sensor-based motion tracking system

(2) Identifying the risk of WMSD

Upper arms, wrists, and hands

(1) 3 I2M IMU Motion tracking system (NexGen Ergonomics Inc, Quebec, Canada, sampling frequency = 128 Hz)

(1) 3 Axes elbow, wrist, forearm, and upper arm joint angles (yaw, pitch, roll; °)

TK Motion Manager/ Human Motion Analyzer

×/×

(4) Akanmu, and Olyawela-2020 [80]

1 Healthy carpenter

Preliminary study

(1) Examining ergonomic exposures of carpentry subtasks that might lead to musculoskeletal injuries

(2) Evaluation of preventive and protective interventions impact on the musculoskeletal injuries

Upper arm, and forearm

(1) 2 smartphones (Sampling frequency = 10Hz) attached with straps and bands

(1) Joint angles (yaw, pitch, roll; °)

(2) Duration of carpentry subtasks (s)

No information

×/✓

(5) Ohlendorf et al. 2020 [55]

20 Dentist teams and a

student control group

Cross- sectional case control

(1) Contributing information on the prevalence of MSD in dental professionals using online questionnaires

(2) Ergonomic risk assessment of treatment procedures based on RULA, focusing on treatment concepts of general dentistry,

endodontology, oral surgery and orthodontics

Shoulders (scapula and humerus), and forearms

(1) 8 IMU sensors embedded in a Lycra body suite (MVN Link, XSens, Enschede, Netherlands, sampling frequency = 240Hz)

(Secondary)

(1) A risk score for every recorded frame using the RULA (a mean score and SD) for wrist and arm

(2) The percentage of spent time in the respective risk scores and at high-risk postures (%)

MATLAB/ Xsens-MVN Analyze

×/✓

(6) Blume et al. 2021 [56]

15 Teams consisting of a dental student a dental assistant trainee

Cross- sectional case series

(1) Investigating the ergonomic risk of

dental students based on RULA score using

inertial sensors during performing standardized dental activities

Upper arms (scapula and humerus), forearms, and wrists

(1) 8 IMU sensors embedded in a Lycra body suite (MVN Link, XSens, Enschede, Netherlands, sampling frequency = 240Hz)

(1) Modified RULA score (1 to 7) and relative time spent at each RULA score

[Using accelerometer, angular velocity, and joint angle values in 3 axes (yaw, pitch, roll; °, °/s, g)]

Xsens-MVN Analyze

×/✓

(7) Maurer-Grubinger et al. 2021 [57]

15 Healthy oral and maxillofacial surgeons

Cross- sectional case series

(1) Developing a script that allows quantification of RULA ergonomic risk in dentistry using IMU measured data of joint angles and positions of body segments

(2) Proposing various modified RULA outcomes

Upper arms (scapula and humerus), forearms, and wrists

(1) 8 IMU sensors embedded in a Lycra body suite (MVN Link, XSens, Enschede, Netherlands, sampling frequency = 240Hz)

(1) Modified RULA score (1 to 7) and relative time spent at each RULA score

[Using accelerometer, angular velocity, and joint angle values in 3 axes (yaw, pitch, roll; °, °/s, g)]

Xsens-MVN Analyze

×/✓

(8) Schall et al. 2021 [60]

35 Subjects [18 manufacturing cyclic workers and

17 non-cyclic workers]

Randomized clinical trial

(1) Comparing mean levels of full-shift exposure summary metrics based on both posture and movement speed between manufacturing cyclic and non-cyclic workers

(2) Exploring the patterns of between- and within-worker exposure variance and between-minute (within-shift) exposure level and variation within each group

Upper arms and dominant wrist

(1) 3 ActiGraph GT9X Link IMUs (ActiGraph, USA, sampling frequency = 100 Hz) fixed with elastic hook-and-loop fastener straps using hypoallergenic cohesive

bandages

(1) Upper arms posture or angles (°), and relative time spent at neutral and extreme states (%)

(2) Angular movement speed (°/s) and relative time spent at low-speed and high-speed states (%)

(3) Relative time spent at 3 states of low speed, neutral, and low speed-neutral of rest/recovery exposure (%)

ActiGraph Software

×/✓

(9) Merino et al. 2018 [63]

3 Workers with musculoskeletal complaints

Cross-sectional case series

(1) Evaluating the risk of musculoskeletal injuries in banana harvesting task using EMG, IMU sensors and subjective outcomes

IMU: upper arms, forearms, and wrists

EMG: upper arms

(1) IMU: 6 Xsens MVN Biomech™ sensors (Enschede, the Netherlands, Sampling frequency = 120 Hz) mounted on and attached with Velcro straps

(2) EMG: Miotec 4-channel

Miotool 400 (sampling frequency = 2 kHz)

(3) Ag/AgCl surface electrodes of Meditrace®) in a bipolar configuration

(1 cm in diameter and an inter-electrode distance of

2 cm)

(1) Mean joint angles (yaw, pitch, roll; °) and time taken to remove the bunches from the stalk (s)

(2) Maximum voluntary isometric contraction (μV), peak muscle use (RMS and percentage) and mean value of signal and median frequency in each muscle (μV and Hz)

Excel/ Xsens MVN Studio Pro/ Miograph

×/✓

(10) Vignais et al. 2017 [76]

5 Healthy workers

Cross-sectional case series

(1) Performing a real-time ergonomic analysis on workers dealing with material handling tasks using IMUs and electro goniometers based on combining RULA computation and a subtask video analysis

(1) IMU: upper arms and forearms

(2) Electro goniometer: wrists

(1) 4 CAPTIV Motion

IMUs (TEA, Nancy, France) fixed with adjustable strap (sampling frequency = 64 Hz)

(2) 2 bi-axial electro goniometers (Biometrics Ltd.,UK) attached by medical tape and straps (sampling frequency = 32 Hz)

(1) RULA score for each joint and an overall score (1 to 7), and percentage of time spent of each joint at risky levels (%)

[Using 3 axes shoulder, elbow, and wrist joint angles (yaw, pitch, roll; °)]

CAPTIV software

×/✓

(11) Vignais et al. 2013 [79]

12 Healthy workers

Randomized clinical trial

(1) Performing a real-time ergonomic analysis of manual tasks in an industrial environment dealing with manual handling tasks using IMUs and electro goniometers based on RULA computation for body overall and local parts

(2) Assessment of an auditory feedback to prevent development of MSK disorders in workers

Similar to previous study

(1) 4 wireless Colibri IMUs (Trivisio GmbH, Trier, Germany, sampling frequency = 100 Hz) wrapped around limbs with adhesive bands

(2) Goniometer type is similar to previous study

(1) RULA score for each joint and an overall score (1 to 7), and percentage of time spent of each joint at risky levels (%)

[Using 3 axes shoulder, elbow, and wrist joint angles (yaw, pitch, roll; °)]

No information

✓/✓

(12) Zhang et al. 2022 [61]

30 Manufacturing workers (2 cyclic/ non-cyclic groups of 15 workers)

Cross-sectional case control

(1) Quantification of self-reported daily discomfort, distraction and burden caused by putting on wearable inertial sensors in manufacturing workers

(2) Evaluating contribution level of different personal and work characteristics on the discomfort, distraction, and burden ratings

Upper arms, and dominant wrist

(1) 3 IMU sensors of ActiGraph GT9X Links (ActiGraph, USA) fixated with elastic hook and loop fastener straps

(1) Upper arm elevation angle (°), and the magnitude of the elevation speeds (°/s)

ActiLife

✓/×

(13) Seidel et al. 2021 [70]

500 Workers

Cross-sectional case series

(1) Providing direct measurements for assessment of workloads of the hand or elbow in work-field based on the Threshold Limit Value (TLV) for Hand Activity Level (HAL)

(2) Finding associations between measured TLV for HAL and disorders or complaints

IMU: shoulders, forearms, elbows, and wrists

EMG: Forearms

(1) CUELA multi-sensor system (IFA, Germany) containing potentiometers and IMUs (Analog Devices ADXL 103/203 3D accelerometers and muRata ENC-03R gyroscopes and goniometers), all embedded in a wearable body-shaped cloth

(2) A 4-channel surface EMG module (BioMed, Germany)

(1) Mean power frequency of the power spectra of angular data (Hz) [Using 3 axes angular values of measured joints (yaw, pitch, roll; °)]

(2) Mean angular velocity (°/s)

(3) Kinematic micro-pauses (%)

(4) HAL exposure categories (Low/Medium/High)

[Using RMS values of EMG signal (V)]

CUELA designed software

×/×

(14) Poitras et al. 2020 [69]

16 Healthy subjects

Cross-sectional case series

(1) Assessment of the concurrent validity of IMU units of MVN, Xsens in comparison to VICON motion capture system during simple tasks and complex lifting tasks

(2) Establishment of the discriminative validity of a wireless EMG system for the evaluation of muscle activity

IMU and markers: shoulders (scapula humerus), and wrists

EMG: shoulders

(1) 6 IMU sensors of MTw (MVN, Xsens Technologies, Enschede, Netherlands) fixed with hook and loop straps around arms

(2) 9 Vicon MX cameras (Vicon Motion Systems Limited, Oxford, UK)

(3) EMG sensors of Trigno Wireless EMG system (Delsys, Boston, MA,

USA)

(1) Shoulder ROM and RMSE value with respect to VICON measurement (°)

(2) RMS EMG (V, and (% of MVC)

Nexus/MVN studio/ MATLAB

×/✓

(15) Bassani et al. 2021 [71]

1 Healthy subject

Preliminary study

(1) Proposing a wearable monitoring system for sEMG and coherent motion data aimed at real-time tracking of workers’ activity for the analysis and prevention of WMSK disorders

IMU: upper arm, forearm, and hand

EMG: forearm

(1) 3 MPU-9250 IMUs (Invensense Inc., USA), enabling 9-axes inertial sensing in addition to a thermal sensor to ease the compensation of

the gyro values all embedded in a package and wrapped around the limbs with a stretchable strap (sampling frequency = 100 Hz)

(2) 8 sEMG electrodes and acquisition systems (proposed system and g.®USBAmp of Guger Technologies)

(3) STM32F407VG ARM Cortex-M4 CPU

(STMicroelectronics)

(1) Signal to noise ratio of EMG signals (SNR)

[Using RMS EMG signal values (mV)]

(2) Acceleration, angular velocity, and angle of joints in all 3 axes (yaw, pitch, roll; g, °/s, and °)

MATLAB

×/✓

(16) Lee et al. 2019 [81]

3 Healthy subjects

Cross-sectional case series

(1) Assessment of the reliability and validity of a novel posture matching method in construction activities using IMU measurements of joint angles

Dominant upper arm wrist, and hand

(1) 3 IMU sensors (I2M IMU system; NexGen Ergonomics Inc., Canada) fixed with stretchable bands and straps through a designed platform

(1) Upper arm joint angles and RMSE values in all 3 axes (yaw, pitch, roll; °)

TK Motion Manager software/ HM Analyzer

×/×

(17) Peppoloni et al. 2016 [72]

10 Healthy supermarket cashiers

Cross-sectional case series

(1) Proposing a novel wearable wireless system capable of assessing the muscular efforts using sEMG and postures of the human upper limb using IMU sensor for WMSK disorders diagnosis

IMU: upper arm, forearm, and hand,

EMG: forearm

(1) 3 IMU sensors embedded inside elastic band (no further information)

(2) 8-Channel sEMG (no further information)

(1) Shoulder, elbow and wrist extension/flexion, and ulnar deviation (yaw, pitch, roll; °)

(2) RMS values of EMG signal (mV) and power spectral density (W/Hz)

(3) Strain inde and RULA score (1 to 7)

Self-developed GUI by MATLAB

×/✓

(18) Battini et al. 2014 [82]

No information

Preliminary study

(1) Introducing an innovative full-body real-time ergonomics assessment system for manual material handling in warehouse environments, using inertial sensors

Shoulders (scapula and humerus),

forearms, and hands

(1) IGS-180i (Animazoo, UK) containing 8 IMU sensors for upper body section and mounted on a light full-body suit (sampling frequency = 500 Hz)

(1) Ergonomic evaluation scores such as RULA, and Lifting Index [Using shoulders (scapula), upper arms, forearms, and hands joint angles and postures with 6-DoF (yaw, pitch, roll; °)

Self-developed software

×/✓

(19) Slade et al. 2021 [83]

5 Healthy subjects

Cross-sectional case series

(1) Presenting the

OpenSenseRT, an open-source and wearable system providing upper and lower extremity kinematics in real time using IMUs

(2) Assessment the accuracy of the proposed system by comparing it to an optical motion capture as the gold standard seeking to achieve an RMSE of 5° or less for 3D joint angles

IMU and markers: Upper arms, forearms, and hands

(1) 6 IMU (ISM330DHCX breakout boards, Adafruit Industries Inc., USA) mounted on stretchable straps

(2) A Raspberry Pi 4b + (Raspberry Pi Foundation) as the microcontroller

(3) Optical motion capture system (Optitrack)

(1) Shoulder joint angles and RMSE during flexion, adduction, and rotation (yaw, pitch, roll; °)

(2) Elbow joint angle and RMSE during flexion (pitch, °)

(3) Wrist joint angle and RMSE during flexion (pitch, °)

Self-developed software by OpenSense tools

×/×

(20) Yang et al. 2020 [58]

116 Surgery cases ( by 53 surgeons)

Cross-sectional case series

(1) Identifying risk factors and assessment of intraoperative physical stressors by subjective (questionnaires) and objective outcomes (IMU measurements)

Upper arms

(1) 2 IMU (APDM Inc, Portland, USA) [model has not been specified]

(Sampling frequency = 128 Hz)

(1) Mean deviation angle of upper arms (°)

(2) Percentage of spent time in the demanding posture during the surgical time (%)

MATLAB

×/×

(21) Hallbeck et al. 2020 [59]

4 Healthy breast surgeons

Cross-sectional case series

(1) Comparing the ergonomics for the beast surgeons between skin-sparing mastectomy (SSM) and nipple-sparing mastectomy (NSM) using subjective and objective measures

Upper arms

(1) 2 IMU (APDM Inc, Portland, USA) [model has not been specified]

(Sampling frequency = 128 Hz)

(Secondary)

(1) Orientation (angles) of upper arms using mean angle (yaw, pitch, roll; °)

(2) Spent time in each RULA level for postures during surgery (%)

MATLAB

×/×

(22) Nath et al. 2018 [84]

2 Healthy workers

Cross-sectional case series

(1) Proposing a new methodology for evaluation of the ergonomic risk levels in warehouse operations caused by overexertion using body motion data

(2) Investigating the appropriate data acquisition and processing settings through a leave-one-subject-out cross-validation framework

Upper arm

(1) 1 Smartphone (Google Nexus 5X or Google Nexus

6) strapped around the limb

(1) Acceleration (g), linear acceleration), and angular velocity in all 3 axes (yaw, pitch, roll; g, m/s², and °/s)

(2) Activity duration (s) and frequency (%)

MATLAB

×/✓

(23) Jahanbanifar and Akhavian-2018 [77]

1 Healthy construction worker

Preliminary study

(1) Developing a framework for quantification of human force as a risk indicator associated with WMSK disorders in construction workers using wearable sensors

Upper arm

(1) 1 Smartphone (sampling frequency = 35 Hz) fixated with a sports armband

(2) SIMULATOR II; Functional Upper Extremity Rehabilitation device (BTEtechnologies inc., USA)

(1) F as the net force exerted (N or kg.m/s²)

(2) P as the power (W or kg.m²/s³)

(3) d as the displacement of the subject’s arm (m)

[Using acceleration, angular velocity and posture values in all 3 axes (yaw, pitch, roll; g, °/s, and °)]

(4) t as the duration of the experiments (s)

Sensor Log smartphone application/ Self-developed Python software

×/✓

(24) Cerqueira et al. 2020 [74]

5 Healthy subjects

Cross-sectional case series

(1) The design and development of an innovative smart garment providing (a) real-time ergonomic risk assessment, (b) objective data measurements to ergonomists, (c) posture awareness to operators through haptic feedback

Upper arms

(1) 2 IMUs; 9250 (Invensense Inc., USA) attached to a skinny fitted shirt (sampling frequency = 100 Hz)

(2) 2 haptic motors; ERM (Precision MicrodrivesTM, London) operating at 80 Hz and 250 Hz

(3) the STM32F4 ARM microcontroller

(1) Upper arm posture, and RMSE in pitch and roll axes (°)

(2) Time spent percentage during each risk state or posture state (%)

MATLAB

×/✓

(25) Singh et al. 2017 [85]

4 Healthy surgeons (gynecologists)

Randomized crossover study

(1) Comparing the effect of

different chairs on WMSD for surgeons during vaginal procedures based on subjective and objective outcomes

(2) Assessment of the WMSD risk in gynecologists

Upper arms

(1) 2 IMU sensors of OPAL(12M SXT version APDM, Inc, Portland,

USA) wrapped around limbs with adhesive bands

(Secondary)

(1) The percentage of time spent in each RULA score for each body part (%) [Using shoulder elevation angle (°)]

No information

×/✓

(26) Lind et al. 2020 [62]

16 Manufacturing plant workers

Cross-sectional case series

(1) Proposing the developed haptic feedback module of the Smart Workwear System platform

(2) Evaluating its user experience and preventive application for reducing biomechanical loads in light repetitive manual tasks

IMU: upper arms

Haptic vibrator: 5cm lower than IMUs

The Smart Workwear System haptic feedback module embedded in a workwear shirt containing:

(1) 2 IMUs (LPMS-B2 IMU, LP Research, Tokyo, Japan, sampling rate = 25Hz) embedded on stretchy workwear shirt

(2) A vibration actuation unit (Precision Microdrives Limited, London, UK, feedback level = 10Hz)

(3) STM32 microcontroller

(1) Angles of upper arm elevation (°)

(2) Proportion of time-spent at upper-arm elevations ≥ 30◦, ≥ 45◦

and ≥ 60◦ (%)

ErgoRiskLogger smartphone application

✓/✓

(27) Granzow et al. 2017 [64]

14 Reforestation hand planters

Cross-sectional case series

(1) Characterizing the trunk and upper arm postures, movement velocities, and neck/shoulder muscle activation patterns in hand planters during full-shift work

(2) Comparing the results with previous findings to

understand the exposures to physical risk factors of hand planters

Shoulder

(1) IMUs: 1 ActiGraph GT9X (Actigraph, USA, sampling frequency = 100 Hz) attached with elastic neoprene straps

(2) EMG electrodes (model SX230, Biometrics Ltd, Gwent, UK)

(1) Shoulder muscle forces as a percentage of

the RMS EMG amplitudes observed for the submaximal reference

contractions (%)

(2) Upper arm flexion/extension angles (°) and angular velocity (°/s)

(3) The ratio of time spent at three postures of neutral, rest and extreme, and at two velocities of low and high (%)

Self-developed GUI by LabView

×/×

(28) Khalil et al. 2021 [65]

34 Baseball pitchers

Cross-sectional case series

(1) Determining the relation between medial elbow torque measured by wearable sensor IMU, and adaptations of the medial elbow structures obtained by dynamic ultrasound imaging in asymptomatic collegiate pitchers

Elbow

(1) 1 Motus Global mobile IMU sensor (Motus Global Inc., USA) embedded in a pocket of a wearable sensor baseball sleeve (Motus Global, Rockville Centre, NY)

(1) Medial elbow torque (N.m)

(2) Arm rotation (maximum angle of the forearm; °),

(3) Arm slot (angle of the forearm in relation to the ground at ball release; °)

(4) Arm speed (maximum rotational velocity of the forearm; rotations/minute)

Motus Global smartphone application (motusTHROW)

×/✓

(29) Villalobos and Maccowlry 2021 [86]

20 Meat cutters

Cross-sectional case series

(1) Presenting an application of IMUs to perform task classification and measure work-related musculoskeletal disorders risks in meat cutters, using artificial intelligence and machine learning techniques

(2) Validation of the proposed application

Wrist

(1) A Wit Motion

BWT901CL Bluetooth 2.0 IMU sensor (Wit intelligence, China) attached with a Velcro strap

(1) Angle, angular velocity, and acceleration of wrist/hands (yaw, pitch, roll; °, °/s, g)

(2) Wrist/hands RULA score (1 to 4) and the spent time in a risky posture based on RULA score (s)

Self-developed GUI

×/✓

(30) Forsman et al. 2021 [78]

6 Healthy subjects

Cross-sectional case series

(1) Developing and testing the validity of a method for

workplace wrist velocity measurements

Wrist and hand

(1) 2 IMUs (Movesense.com, Suunto, Vantaa, Finland, sampling frequency = 50 Hz) attached with double-sided tape or an armband

(1) Wrist angular velocity (yaw, pitch, roll; °/s)

Self-developed smartphones app

×/✓

(31) Rodríguez-Vega et al. 2022 [73]

1 Healthy subject

Preliminary study

(1) Development of measurement and classification of hand movements at work, using a hand-motion capture system

Hand and fingers

(1) 6 IMU sensors

(2) 6 Force-resistive sensors (No further details)

(1) Triaxial acceleration, angular velocity, and magnetic field (yaw, pitch, roll; m/s2, rad/s, and µT)

(2) Exerted force

each fingertip (μV)

MATLAB

×/×