Ergonomics evaluation of lawn mower operator’s working posture using JACK software and kinect interface

Abstract

BACKGROUND:

Nowadays, real-time motion tracking devices are widely used for ergonomic assessment of several manual quotidian activities. The real-time tracking of human activities makes it easier to observe the exposure of work-related musculoskeletal disorders (WMSDs) in the human body.

OBJECTIVE:

This study aims to determine the suitability of a real-time motion tracking device (Kinect v1 interfaced with a commercial ergonomic assessment software, JACK) for real-time ergonomic evaluation of the strenuous operation of the manual lawn mower.

METHOD:

The lawn mower operators perform various strenuous activities while operating the manual lawn mower for long intervals of time, which causes WMSDs in the entire body of the operators. These working operators’ activities have been captured using Kinect v1 interfaced with JACK, to address the ergonomic issues responsible for the whole-body WMSDs. The forces acting on the lower back, Rapid Upper Limb Assessment score and static strength have been predicted using JACK.

RESULTS:

This study proves the exposure of the operators towards the whole-body WMSDs while operating the manual lawn mower.

CONCLUSION:

The findings provide a quick and straightforward approach for performing the real-time ergonomic evaluation of any operation, which can help the industrial staff estimate the risk of level WMSDs.

Keywords

1 Introduction

In recent years, ergonomic aspects are playing a vital role in the better designing of workspace with comfortable and safe working conditions for the operators. Ergonomically designed systems and products are the right fit for the people and can be used less physically and mentally. This improves performance and reduces the fatigue level of the operators. Ergonomic strategies are broadly implemented to provide safe working conditions for the operators. However, the sixth European working conditions survey proves the existence of repetitive movements, awkward postures, excess fatigue and insufficient strength for performing the numerous quotidian tasks [1]. These awkward postures and fatiguing conditions affect operators’ health and could cause work-related musculoskeletal disorders (WMSDs). WMSDs are commonly occurred in the upper limb, knee, neck and lower back of the working operators [2]. The WMSDs play a vital role in causing occupational diseases, operator disability and frequent absenteeism [3, 4]. In fact, WMSDs are the primary cause of health issues and disabilities. It occurs because of various physical and psychosocial risk issues, and inappropriate design of the working process and system [5]. It was observed that there is a strict relation between workplace design, repetitive activities and ergonomics concerns, with the existence of WMSDs and operators efficiency [6]. A poor workplace and equipment design provides vulnerable working conditions that compelled the operators to perform the assigned tasks by maintaining awkward postures for a prolonged period [7]. Supplementary, these working conditions cause fatigue and disorders in the lower back, knees, shoulders and arms of the operators [8]. The WMSDs were identified with expert system Sonex which provides fast results [9].

In routine days, the operation of a manual lawn mower is commonly observed in our surroundings. The manual lawn mower operators experience neck pain, lower back pain, knee pain, leg fatigue, and feet discomfort. The lawn mower operators are vulnerable to the awkward postures and large operating forces, leading to the upper body injuries and WMSDs [10]. The dynamic pulling forces critically affect the shoulder and upper arm of the operators while operating the lawn mower. However, the manual lawn mowers are still crucial in this era of automation due to its lesser purchasing cost, low maintenance and more reliability than that of the automatic lawn mowers. It is necessary to develop an effective method for proper evaluation of riding lawn equipment to improve operators’ efficiency and safety by minimizing the risks of WMSDs [11 –13]. Therefore, this operation is selected in the present study for performing ergonomics assessment. To minimize the risk of WMSDs, evaluating risk factors at the workplace guided the implementation of ergonomic interventions to redesign the workplace [14, 15]. WMSDs can be prevented by ergonomically analyzing the working systems’ design in manufacturing and service industries [16 –18]. The Rapid Entire Body Assessment (REBA) identified insensitive postures and applied regression analysis to determine the significance of weights contribution to the final REBA [19]. Fuzzy sets utilized for REBA scoring system [20]. Many researchers have been widely evaluated the manual activities and working environment using various observational approaches such as OWAS (Ovako Working posture Assessment System), NIOSH (The National Institute for Occupational Safety and Health), RULA (Rapid Upper Limb Assessment), OCRA (Occupational Repetitive Actions), REBA and LUBA (Loading on the Upper Body Assessment) to reduce the risk of WMSDs [21 –23]. These approaches require an ergonomics expert and consume a lot of time to analyse a worker’s posture. The real-time motion tracking devices were introduced mainly for quick data collection and accurate real-time ergonomic evaluation. These devices help implement the anthropometric and biomechanical approaches simultaneously for ergonomic analysis to redesign and modify the existing workplace [24].

Microsoft Kinect v1 was used for real-time data collection, and it requires lesser time for data collection with higher accuracy [25, 26]. Initially, Microsoft Kinect v1 was introduced as a gaming device for Xbox 360 [27]. A webcam and add-on module of the Kinect v1, enables the users to control the computer by making a natural user interface using voice instructions and gestures. Nowadays, Microsoft Kinect v1 being used in the various research applications, because of its high accuracy and quick operation time [28 –30]. The Kinect v1 is proficient enough to capture the human activities with accurate joint positions displayed in a line diagram of the human skeleton [31]. The Kinect v1 can be integrated with JACK to form an interface for a real-time ergonomic evaluation. JACK is a commercial Digital Human Modelling (DHM) software, consists of a task analysis tool (TAT) kit module to evaluate the postural risk of human activities in real-time [32, 33]. JACK provides a set of tools for ergonomic analysis to precisely assess the manual activities and design the workplace accordingly [34]. These tools include RULA, lower back analysis (LBA), static strength prediction (SSP), metabolic energy expenditure, material handling limits, NIOSH, OWAS and fatigue recovery The operator’s actual activities can be imported in the JACK to develop a digital human model with a similar virtual environment to perform an ergonomic analysis [35].

Kinect v1 and JACK’s combined interface can perform the real-time ergonomic evaluation in various fields such as healthcare, industries, education, surveillance, and sports. Martin et al. [36] carried out real-time ergonomic analysis using a single Kinect for lifting operation and recommended the same can be useful for the training program. However, Kinect v1 supports the various ergonomic observational techniques to evaluate the input parameters in different loading conditions [37]. The Kinect is capable of capturing motion with some minor adjustments [38]. According to the researchers, the real motion tracking capabilities of Kinect becomes a useful input to the JACK for better ergonomic analysis. This study’s objective is to experience JACK and Kinect v1 interface capabilities to perform a real-time ergonomics evolution of the operators while operating lawn mower. A human performs many strenuous daily routine activities by maintaining awkward postures without being aware of its consequences. Most of these activities require a massive push/pull forces, and among these activities, the operation of manual lawn mower is considered for the present study. The manual lawn mower is used for grass cutting and smoothing the lawns, which requires a consistent push/pull force during operation. The ergonomics approaches such as RULA, LBA and SSP, have been used to analyze the operators’ postures while operating the manual lawn mower. Experiments have been conducted in a lawn. The static strength, forces acting on the lower back and RULA score are predicted by using an interface of JACK and Kinect v1. Precisely, the compression and shear forces acting on the lower back’s L4/L5 spinal segment are measured. The interface of JACK and Kinect v1 can be used as a practical approach to performing the real-time ergonomic evaluation. This approach requires less time for execution. It partially eliminates the presence of ergonomics experts for performing the real-time ergonomic evaluation. After the initial training, even the non expert can use this approach for the initial assessment of WMSDs.

In the following section, the RULA, LBA and SSP tools are briefly explained. A systematic experimental procedure followed to collect the operators’ data, using Kinect v1, is described in the next section. The collected data are further processed and analyzed in JACK with the help of above-listed ergonomic approaches. The results obtained by all three techniques for each operator are discussed in the further section. The successive section represents the discussion about the feasibility and implications of the present study. Furthermore, the concluding points are summarized in the last section.

2 Materials and methods

2.1 Rapid Upper Limb Assessment (RULA)

RULA is one of the most popular ergonomic assessment tools for posture analysis. RULA is used to quantify the efficacy of Ergonomics and biomechanics. An ergonomics assessment performed by implementing RULA to examine the working posture of robotic surgeons. RULA with real-time sensing and skeleton tracking system provides more accurate results [39]. RULA’s useful scoring system allows you to take a snapshot of the highest risk posture adopted during the task. The scoring system is broken down into four actions levels with indications in as to the urgency of the investigation.

Action level 1. Score of 1–2 = Acceptable

Action level 2. Score of 3–4 = Investigate further

Action level 3. Score of 5–6 = Investigate further and change soon

Action level 4. Score of 7 = Investigate further and change immediately

RULA is designed to assess the force, posture and movement associated with sedentary tasks including manufacturing, retail, computer tasks, laboratory work or where the individual is seated or standing without moving about.

A framework described by combining the Kinect v1 with the RULA method for 3D motion analysis [40]. Kinect used to develop a semi-automatic RULA for the real-time evaluation of upper body movements [41]. A semi-automatic interface named K2RULA created using Kinect to perform real-time and offline RULA [15]. Kinect v1 is used to capture the working operator’s various awkward postures and import the JACK software’s captured data as an input to calculate the RULA score. Wherein, the posture of the operator is divided into two groups: body group A posture and body group B posture. The body group A posture includes upper arm, lower arm, wrist, and wrist twist of the operator, whereas body group B posture has its neck and trunk. Some inputs such as frequency of the tasks, arms support, working position and the load on the operator’s limbs are entered manually in the JACK. JACK considers all provided inputs and generates a grand score. The grand score indicates the following actions required to minimize the risk of upper body WMSDs:

Grand score: 1–2 directs that the working posture is acceptable if it is not maintained for a prolonged time interval.

Grand score: 3–4 indicates that a further investigation is required, and changes may be needed.

Grand score: 5–6 indicates that an investigation and changes are required soon.

Grand score: 7 indicates that an investigation and modifications are needed immediately.

RULA appoints risk factors for a specific manual operation following a person’s body posture defined in JACK’s environment. The repetitiveness of operation, weight of the load and the duration for which the load is held, are provided as input to JACK for defining the body posture. The joint angles and twisting of the legs, trunks arms, neck and wrist, are continuously capturing by the Kinect v1 and provide to the JACK for performing RULA. Furthermore, the JACK analyses the most awkward posture based on the data provided by Kinect and manual inputs such as weight of load, duration, and frequency of the operation. Moreover, RULA is also required to specify that the load is static or dynamic. Herein, the lawn mower operators’ motion is captured by Kinect v1 and then, processed in JACK software to generate RULA grand scores. JACK generates RULA grand scores for the most awkward posture of the working operator.

2.2 Static Strength Prediction (SSP)

The SSP is an ergonomics approach, provided in the JACK for predicting the strength required to perform an assigned activity. SSP approach can also calculate the operator’s torques and joint angles while working in an awkward posture. The joint angles include limb angles for the knee, wrist, elbow, shoulder, ankle and hip and, trunk angles for lateral bending flexion and rotation. The torque and muscle effect, including flexion, abduction, extension and adduction, are also measured. SSP predicts the required strength by analyzing the cumulative impact of working posture, anthropometric characteristics and loads on the operator’s hands. SSP approach indicates the percentage of male or female operators having enough strength to perform a particular task [42]. This approach’s functionality is based on the University of Michigan’s 3D Static Strength Prediction Program (3DSSPP). SSP approach has capabilities to analyze operators’ slow movements by neglecting the effects of acceleration and momentum. SSP tool also illustrates the measured torques, joint angles and percentage of operators expected to have required strength, in the form of a graph. The SSP results help design or redesign the safe working environment following the operators’ strength capabilities. Hovanec [43] highlighted Tecnomatix JACK’s use to perform various ergonomic analyses in the field of human factor reliability. SSP measures the torque, joint angles, and the percentage of operators having enough to perform an assigned operation, following the operators’ most awkward posture, anthropometry, and gender. It also considers the hand tools and nature of applied loads on the hands of the operator. In the present study, the lawn mower operators’ body movements are captured by Kinect v1 and then processed in JACK software to predict the operators’ static strength. The results of the Static Strength Prediction (SSP) signify the comfortably of the lawn mower operation performed by the operators.

2.3 Lower Back Analysis (LBA)

LBA tool is provided in the JACK for measuring the forces acting on the lower back of the working operator. The measurement of the forces is based on a complex biomechanical model [44]. The real-time motion is captured by Kinect v1 and imported into TAT of JACK for performing LBA. The LBA tool measures the compression, Anterior/Posterior (AP) shear and lateral shear forces acting on the operator’s L4/L5 spinal segment. This measurement of lower back forces indicates the risk of WMSDs to the operators. A study was performed using LBA in JACK to calculate the forces acting on the operator’s lower back while driving a lift truck [45]. Kinect v1 provides the working operator’s real-time motion to the manikin designed in JACK’s environment for defining the actual body posture.

Furthermore, the load on the hands of the operator is specified. LBA tool compares the calculated compression forces with permissible limit recommended by NIOSH (National Institute for Occupational Safety and Health). This comparison represents in the form of graph including yellow, green and red colour codes. Yellow colour indicates that the calculated compression forces are less than in permissible limit of NIOSH and no changes required in the working environment. Green colour signifies that calculated compression forces are still less than the allowable limit of NIOSH. It may even cause WMSDs if the same working conditions are maintained for a prolonged period and recommend changes. Red colour indicates that the calculated compression forces are higher than the permissible limit of NIOSH, and immediate changes are necessary for the working environment. The LBA tool provides one more graph representing various torques such as sagittal, lateral and axial spinal reaction moments on the L4/L5 spinal segment for signifying the influence of hand loads and weight of the upper limb. This also includes torso muscle tensions such as erector spinae, internal and external obliques, latissimus dorsi, etc. In this study, the lawn mower operators’ working activities are captured by Kinect v1 and then, processed in JACK software for measuring the forces acting on the lower back of the working operator. The Lower Back Analysis (LBA) results estimate the compression, Anterior/Posterior (AP) shear and lateral shear forces acting on the L4/L5 spinal segment of the operator operating the lawn mower. Experimentation design.

The present study is performed in the lawn with average grass height. The grass’s average size is approximately 8.89 centimeters similar in the area wherein three operators work. Three operators from the gardening staff participate in this study for working on the same area of grass. All the participants are observed keenly to assure they are normal health before conducting the study. Kinect v1-JACK interface is used for performing the ergonomic analysis of the operators while operating the manual lawn mower. Kinect v1 collects the operator’s real-time motion and varying posture and sends to JACK as an input. JACK software develops a virtual environment using real-time data Kinect v1, for ergonomically analyzing the operator’s posture.

2.4 Development of human model using JACK

All three skilled operators belong to the age group of 21–30 who operate the lawn mower every day. The stature and weight of operators are measured following the standard procedure, as shown in Table 1. The measured body dimensions are entered into the build human tab of JACK software to create the manikins of similar body features as of real operators. JACK consists of different percentiles for the defined population and race for the manikin development, following each continent’s database. In the present study, 50th percentile of Asian Indian database is adopted to build manikins for the ergonomics analysis. The anthropometric data, along with all inputs, is partially entered in JACK to build a manikin for operator-1. Similarly, the manikins are created to replace operator-2 and operator-3 in the simulated environment of JACK

Table 1
Anthropometric characteristics of the operators

Name Gender Age Standing height (cm) Body-weight (Kg)

Operator-1 Male 21 166.40 63.5

Operator-2 Male 25 175.49 77.9

Operator-3 Male 30 176.00 79.3

Name	Gender	Age	Standing height (cm)	Body-weight (Kg)
Operator-1	Male	21	166.40	63.5
Operator-2	Male	25	175.49	77.9
Operator-3	Male	30	176.00	79.3

2.5 Experimental procedure

Kinect v1 consists of an RGB camera that can detect faces by capturing the three colour scenes (red, green and blue). A depth sensor is used to records 3-dimensional space. A multi-array microphone used for voice commands and a tilt motor utilized to facilitates the tilting of a camera following the region of view. Accordingly, the video and depth sensor can run at 30 frames per second with 640^*480-pixel resolution. The parts and coordinates of the depth sensor of the Kinect v1 are shown in Fig. 1.

Fig. 1

Kinect v1 with its parts and coordinates of the depth sensor.

Kinect v1 is placed at the height of 1 meter from the ground, and about 1.7-metre space is provided between Kinect and working operator. The position of the Kinect v1 is focused on capturing the complete activity performed in the surrounding environment. Kinect v1 is plugged with the JACK and trigger a module built-in in JACK, to activate the Kinect-JACK interface. Kinect v1 follows the motion of the operator and update the camera and JACK window accordingly. Furthermore, Kinect v1 records the operator’s movements and tracks the joint information by converting the captured human into a skeleton line diagram. Kinect v1 captures the movements associated with the left portion of the operators. The movements of the operator captured by Kinect v1, are displayed in the environment of the JACK. The motion capturing of the operators is captured in the working area covered by the curtains. The present study is performed in an ideal environment of a lawn during clear weather conditions, making it simple for the Kinect v1 to capture the operator’s body movements.

Participating operators follow some instructions such as working in the reach of the Kinect v1 to standardize the study. Numerous factors, such as the position of Kinect and object, type of operation, size of object and weather conditions, may influence the motion capturing feasibility of Kinect v1. The inadequate factors may result in errors in the research. Therefore, the present study is performed in clear weather conditions. For concerning the participating operators’ privacy, the analysis is performed when the participating operators wear full cloths.

The real-time data of the operators has been recorded for 30 Seconds to 45 Minutes. Kinect v1 captures all real operators’ movements and updates manikin’s motion following the changes in the real operator’s movements. The captured data are continuously monitored and analyzed using ergonomic approaches such as RULA, LBA and SSP. The ergonomic analysis is performed by capturing the operators’ postures and displayed in the JACK window, as shown in Fig. 2. The real-time data collected by Kinect v1 as shown in Fig. 3a and, this data is used as input to perform RULA as illustrated in the Fig. 3d. Moreover, the SSP and LBA are performed in the environment of JACK as shown in Fig. 3c and 3b, respectively.

Fig. 2

Ergonomic analysis by tracking the motion of the operator while using a manual lawn mower: (a) Kinect-JACK interface screen for motion tracking; (b) calculation of RULA score for posture analysis of the operator; (c) LBA of the working operator; (d) SSP for the working operator.

Fig. 3

RULA scores for the operators while operating a manual lawn-mower: (a) for operator-1; (b) for operator-2; (c) for operator-3.

3 Results and discussion

In the real-time ergonomic analysis, the operators’ movements are continuously captured while operating the manual lawn mower. Three experiments are performed for the real-time ergonomic evaluation of the manual lawn mower operation. The complete operation of the operator is analyzed. The RULA score values, static strength, and lower back forces are continuously varying following the operator’s posture variations. The Kinect v1-JACK interface generates results for the most awkward pose, which indicates the symptoms of excessive fatigue and WMSDs. The high values of RULA score are predicted for the most awkward posture and also validated with the critical observation of the SSP. The operators’ body movements are captured by Kinect v1 and processed in JACK to obtain the RULA, SSP, and LBA result. The RULA scores, static strength values, and forces acting on the operator’s L4/L5 spinal segment are continuously varied while using the lawn mower. The RULA, SSP, and LBA for all experiments are performed by capturing each operator’s most awkward postures and described in the following subsections.

3.1 RULA grand scores for each operator

The operators’ body movements are captured by Kinect v1 and processed in JACK to generate RULA grand scores. The RULA scores are varied continuously while the operator is operating the lawn mower. JACK generates RULA grand scores for the most awkward posture of the operator. RULA grand scores indicate that the existing operation influences the operator’s body parts and causes excess fatigue and WMSDs. Figure 3 represents the RULA scores for each operator. The operation mainly affects the neck and trunk of the operator-1, and the grand score is found 4 and appears in yellow, as shown in Fig. 3(a). This indicates the working posture of operator-1 is the acceptable range if the same posture is not maintained for a prolonged interval of time. However, the neck, lower arm and trunk of operator-2 are affected while operating the manual lawn mower. The grand score for operator-2 is 7 and appears in red colour, as shown in Fig. 3(b). This indicates that working posture is not acceptable, and immediate changes are required in the existing operation.

Similarly, the neck, lower arm and trunk of operator-3 are affected while operating the lawn mower. For the assessment of operator-3, the grand score is observed as 7 and appears in red colour, as shown in Fig. 3(c). The RULA scores indicate that the existing lawn mower operation and current working posture are unacceptable for operator-2 and operator-3, whereas it seems acceptable for operator-1. However, the operational posture of the operator-1 may also cause WMSDs due to working in the same posture for a prolonged period. Therefore, there are requirements to modify the existing handle design by adjusting the length of handle design to reduce the risk of WMSDs. A well-designed job, supported by a well-designed workplace and proper tools, allows the worker to avoid unnecessary motion of the neck, shoulders and upper limbs. Proper design of lawn mower significantly decreases the force needed to complete the task. The proposed system is found to be more compact, user friendly and less complex, which can readily be used in order to perform several tedious and repetitive tasks.

3.2 Static Strength Prediction (SSP) for each operator

The SSP results for each operator validate the RULA scores as SSP directly influences each operator’s body parts. SSP report for each operator is generated by JACK, as shown in Fig. 4. Green colour signifies that the operator has enough strength for comfortably performing the assigned operation. In contrast, the red colour indicates that the operators lack the strength required to complete the given operation. SSP report represents that some body parts of the operator have insufficient strength to perform an operation. The left shoulder, trunk and hip of operator-1 have insufficient strength wherein left shoulder has 94%, the trunk has 99%, and hip has 98% of the required strength as shown in Table 2. During the operation, the body parts of operator-2 such as left shoulder have 74%, the trunk has 99%, the ankle has 95%, and hip has 97% of the required strength as described in Table 3. Similarly, the SSP indicates that the body parts of the operator-3 such as left shoulder have 89%, the trunk has 99%, ankle 97% and hip has 99% of the required strength as shown in Table 4. Accordingly, this indicates that the changes are needed to exiting a manual lawn mowers exiting design and operation.

Fig.4

Static Strength Prediction for each operator: (a) for operator-1; (b) for operator-2; (c) for operator-3.

Table 2

Static strength summary for the left portion of operator-1

		Operator-1
		Left					Right
		Moment (Nm)	Muscle effect	Mean (Nm)	SD (Nm)	Cap (%)	Moment (Nm)	Muscle effect	Mean (Nm)	SD (Nm)	Cap (%)
Wrist	Flexion/Extension	–0.1	–	–	–	100	0.1	–	–	–	100
	Radial/Ulnar deviation	0	–	–	–	100	–0.1	–	–	–	100
	Sup/Pro	0	–	–	–	100	0	–	–	–	100
Elbow		0.2	–	–	–	100	–1	Flexion	70.4	17.3	100
Shoulder	Abduction/Adduction	–4.7	Abduction	76.8	18.9	100	–5.5	Abduction	72.3	17.8	100
	Rotation backward/forward	–5.5	Forward	95.3	26	100	3.3	Backward	71.6	20.9	100
	Humeral rot	1.5	Medial	2.4	0.6	94	–0.9	Lateral	46.7	10.6	100
Trunk	Flexion/Extension	–54.2	Extension	262.9	82.8	99
	Lateral bending	–35.8	Right	171	37	100
	Rotation	8.6	CCW	102.2	27.3	100
Hip		–33.5	Extension	202.9	81.4	98	–13	Extension	196.3	78.8	99
Knee		4.5	Extension	105.9	37.1	100	–0.8	Flexion	145.3	42.8	100
Ankle		18	Flexion	122.3	40.5	100	3.9	Flexion	163.8	54.2	100

Note: SD = Standard deviation, Cap = Capacity. Sup = Supination, Pro = Pronation.

Table 3

Static strength summary for the left portion of operator-2

		Operator-2
		Left					Right
		Moment (Nm)	Muscle effect	Mean (Nm)	SD (Nm)	Cap (%)	Moment (Nm)	Muscle effect	Mean (Nm)	SD (Nm)	Cap (%)
Wrist	Flexion/Extension	–0.1	–	–	–	100	0	–	–	–	100
	Radial/Ulnar deviation	–0.1	–	–	–	100	–0.2	–	–	–	100
	Sup/Pro	0	–	–	–	100	0	–	–	–	100
Elbow		–1.4	Flexion	67.9	17	100	–2.4	Flexion	75.8	19	100
Shoulder	Abduction/Adduction	–6	Abduction	66.3	16	100	1.6	Adduction	102	32	100
	Rotation backward/forward	2.6	Backward	62.4	18	100	1	Backward	90.2	26	100
	Humeral rot	1.2	Medial	1.5	0.4	74	0.1	–	–	–	100
Trunk	Flexion/Extension	–77.8	Extension	273	86	99
	Lateral bending	–24.8	Right	214	46	100
	Rotation	0.7	CCW	82.2	22	100
Hip		–49.7	Extension	206	83	97	–29	Extension	197	79	98
Knee		–31.2	Flexion	145	43	100	–20.8	Flexion	145	43	100
Ankle		–59.9	Extension	135	45	95	–29.2	Extension	173	57	99

Note: SD = Standard deviation, Cap = Capacity. Sup = Supination, Pro = Pronation.

Table 4

Static strength summary for the left portion of operator-3

		Operator-3
		Left					Right
		Moment (Nm)	Muscle effect	Mean (Nm)	SD (Nm)	Cap (%)	Moment (Nm)	Muscle effect	Mean (Nm)	SD (Nm)	Cap (%)
Wrist	Flexion/Extension	–0.2	–	–	–	100	0.1	–	–	–	100
	Radial/Ulnar deviation	–0.1	–	–	–	100	–0.1	–	–	–	100
	Sup/Pro	0	–	–	–	100	0	–	–	–	100
Elbow		–1.1	Flexion	67.4	16.6	100	–1.4	Flexion	71.7	17.6	100
Shoulder	Abduction/Adduction	–5.6	Abduction	74.6	18.4	100	–4.9	Abduction	68.3	16.8	100
	Rotation backward/forward	–4.6	Forward	97.5	26.6	100	2.8	Backward	79.6	23.2	100
	Humeral rot	1.8	Medial	2.6	0.7	89	–1.3	Lateral	37.6	8.5	100
Trunk	Flexion/Extension	–24.3	Extension	240	75.6	100
	Lateral bending	–69.6	Right	132.6	28.7	99
	Rotation	13.3	CCW	109.4	29.3	100
Hip		–11.5	Extension	203.3	81.6	99	–10.1	Extension	195.8	78.6	99
Knee		–31.8	Flexion	145.3	42.8	100	–21	Flexion	145.3	42.8	100
Ankle		–68.8	Extension	173.2	57.3	97	–52.9	Extension	133.2	44	97

Note: SD = Standard deviation, Cap = Capacity. Sup = Supination, Pro = Pronation.

Each operator’s limb and trunk angles have been calculated while maintaining the most awkward posture during the operation. The trunk angles are computed for lateral bending, flexion and rotation of the trunk. The limb angles are calculated for the wrist, shoulder, knee, hip, humeral rotation, elbow and ankle. A torque is induced on the body limbs and trunk. The limb and trunk angles for operator-1, operator-2 and operator-3 are described in Table 5. The operator capability measurements, limb and trunk angles, induced torque, and population strength means are also represented in the form of a real-time graph with Fig. 5.

Table 5

Limb and trunk angles of the operator-1, operator-2 and operator-3

Limb and trunk angles	Operator-1		Operator-2		Operator-3
	Left	Right	Left	Right	Left	Right
Wrist flexion/extension	0	0	0	0	0	0
Wrist radial/ulnar deviation	0	0	0	0	0	0
Forearm sup/pro	0	0	0	0	0	0
Elbow included	112.8	92.2	62.3	77.7	68	108.8
Shoulder vertical	71.7	77	36.2	17.4	46.5	57.4
Shoulder horizontal	176.5	52	–72.5	92.4	168	81.2
Humeral rotation	–69	–5.4	–5.6	34	–90.6	71.5
Hip included	130.1	159	121.7	154.3	128.9	169.6
Knee included	174.9	174.9	174.9	174.9	174.9	174.9
Ankle included	100.9	73.3	92.5	67.3	67	93.7
Trunk flexion	88.1		81.1		91.1
Trunk lateral bending	–4.3		–1.8		8.8
Trunk rotation	23.6		–7.5		31.5

Note: Sup = Supination, Pro = Pronation.

Fig. 5

Graphical representation of the operator’s capability, limb and trunk angles, induced torque and population strength means for (a) operator-1; (b) operator-2; (c) operator-3.

3.3 Compressive and shear forces acting on the L4/L5 spinal segment

The compression, AP shear and lateral shear forces acting on the L4/L5 spinal segment are calculated for each operator while performing the activity. The compression forces have the highest impact on the L4/L5 spinal segment of each operator. Although, the calculated compression forces are found less than compression force limit recommended by NIOSH as shown in Fig. 6, wherein, (a) LBA for operator-1; (b) LBA for operator-2 and (c) LBA for operator-3. The calculated compression forces for each operator in green colour indicate that the compressions are still less than NIOSH’s permissible limit. But it may cause a nominal risk of lower back injuries and WMSDs for healthy operators as they worked for a prolonged period by maintaining the same posture. Accordingly, the changes are required in the existing design and operation of a manual lawn mower. The distributed moment histogram (DMH) moment distribution and muscle tensions of each operator, including are also represented in the form of graphs, as shown in Fig. 6. The DMH moments include various torques such as sagittal, lateral and axial spinal reaction moments acting on the L4/L5 spinal segment, signifying the influence of hand loads and the upper limb’s weight. Furthermore, the muscle tensions include erector spinae, internal and external obliques and latissimus dorsi, etc.

Fig. 6

(1) Compression and shear forces acting on the L4/L5 spinal segment of each operator: (a) for operator-1; (b) for operator-2; (c) for operator-3; (2) Graphical representation of the DMH moment distribution, muscle tensions, movements of the L4/L5 spinal segment and forces acting on the lower back of the operators: (a) for operator-1; (b) for operator-2; (c) for operator-3.

The present study is focused on understanding and visualizing the feasibility of Kinect v1 to measure various joint angles and capture the awkward posture of working operator. The present study further aimed at understanding the feasibility of JACK software to perform the real-time RULA, LBA and SSP, while interfaced with the Kinect v1. The practical feasibility of the Kinect-JACK interface is observed in the results of this study. This study’s findings make it easier for the industrial and managerial staff to perform an ergonomics analysis of any operation and estimate the risk of level WMSDs. Specifically, this study introduced a real-time ergonomics evaluation using Kinect-JACK interface which consumes less time and eliminates the requirement a highly-skilled ergonomic expert for performing ergonomics analysis. Green and red colours in the SSP reports signify that the operators have or lack the strength required to perform the assigned operation.

Similarly, colour codes of RULA and LBA are representing the risk of WMSDs and lower back injuries. Accordingly, a person unskilled in the art of ergonomics and human factors can evaluate the risk of WMSDs, lower back injuries and excess fatigue. This study provides a quick and straightforward approach for real-time ergonomic evaluation by combining the Kinect v1 and JACK software. From the present study, Kinect and JACK’s interface could be effectively used to measure joint angles and perform real-time ergonomics while the operator is working in a most awkward posture. This interface can also be used in service industries, production industries, and the health care sector to perform the quick ergonomics evaluation. The ergonomic assessment indicates the availability of awkward postures and risk of WMSDs in the body of operators. If the ergonomics evaluation suggests the risk of WMSDs, there seems to be an immense need further to investigate the product and working system’s design.

4 Conclusions

This study indicates that Kinect v1 is a useful tool for tracking the real-time motion of the operators. The interface of JACK and Kinect v1 was found to be a classical approach for the real-time ergonomics evaluation of working operators. The Kinect v1-JACK interface automatically records the different body postures at a high sampling frequency and accurate risk exposure duration. The main contribution of this study is summarized as:

The prediction of static strength and RULA score is found to be critical for the operation of operator-1, operator-2, and operator-3. This indicates that the manual lawn mower’s inadequate design and repetitive execution of awkward postures critically affect the shoulder, torso, knee, ankle, hip and other body parts of the operators.

The forces acting on the L4/L5 spinal segment are acceptable by comparing with the NIOSH standard force limits for the current working conditions. Still, it may cause the risk of the WMSDs and lower back injury if operators are continuously using existing manual lawn mower.

The points mentioned above indicate an immense need to modify the existing manual lawn mower design to reduce the risk of WMSDs and enhance the operators’ safety. Moreover, Kinect v1-JACK interface provides sufficient details to design a new working system or modify an existing working system by considering ergonomics aspects, which can partially eliminate ergonomics experts’ presence and require less time effort performing the real-time ergonomic evaluation. After the initial training, even the operators can use this approach for the initial assessment of WMSDs. Furthermore, some modern types of the lawn mowers such electrically and mechanically powdered, can also be used to ease the task of operators. This study’s findings provide a quick and straightforward approach for the industrial and managerial staff to perform a real-time ergonomic evaluation of any operation and estimate the risk of level WMSDs. The research work can be extended in the evaluation of more complex and repetitive operations performed in daily routine, manufacturing and service industries with consideration of exposure to noise and vibration. It using more than one Kinect device for better motion capturing at the possible directions and orientations.

Footnotes

Acknowledgments

The work was supported by the Centre of Excellence (Industrial and Product Design) of the Department of Production Engineering of PEC Chandigarh, India.

Conflict of interest

The authors declare no potential conflicts of interest with respect to the research, authorship, and publication of this article.

References

Foundation

First findings: Sixth European working conditionssurvey: Résumé. Publications Office. 2015;8.

Abas

, Zamzam

, Mohammad

Work-related Musculoskeletal Disorders (WMSDs) Survey Among Workers at Two Precast Concrete Plants in Johor, Malaysia. International Journal of Academic Research in Business and Social Sciences (2020);10(5), 843–856.

Xia

, Sun

, Chen

, Wang

, Xie

In The Survey of Hand Grinding Operation Work-Related Musculoskeletal Disorders, International Conference on Man-Machine-Environment System Engineering, 2020; Springer:2020; pp 771–777.

Russo

, Di Tecco

, Fontana

, Adamo

, Papale

, Denaro

, Iavicoli

Prevalence of work related musculoskeletal disorders in Italian workers: is there an underestimation of the related occupational risk factors?

BMC Musculoskeletal Disorders (2020);21(1), 1–16.

Merlino

, Rosecrance

, Anton

, Cook

TM.

Symptoms of musculoskeletal disorders among apprentice construction workers. Applied occupational and environmental hygiene (2003);18(1), 57–64.

Grosse

, Dixon

, Neumann

, Glock

CH.

Using qualitative interviewing to examine human factors in warehouse order picking. International Journal of Logistics Systems Management (2016);23(4), 499–518.

Obinna

, Sunday

, Babatunde

Ergonomic assessment and health implications of classroom furniture designs in secondary schools: a case study. Theoretical Issues in Ergonomics Science (2021);22(1), 1–14.

Mavor

, Graham

RB.

The effects of protective footwear on spine control and lifting mechanics. Appl Ergon (2019);76, 122–129.

Pavlovic-Veselinovic

, Hedge

, Veselinovic

An ergonomic expert system for risk assessment of work-related musculo-skeletal disorders. Int. J. Ind. Ergon (2016);53, 130–139.

10.

Lowndes

, Heald

, Hallbeck

MS.

Ergonomics and comfort in lawn mower handle positioning: An evaluation of handle geometry. Appl Ergon (2015);51, 1–8.

11.

Deng

, Shirley

, Rose

, Feltner

, Chen

, Hoyle

, Dutt

, Kaber

Development of a Usability and Functionality Assessment Tool for Riding Lawn Equipment. Proceedings of the Human Factors and Ergonomics Society Annual Meeting (2017);61(1), 2015–2019.

12.

Ren

, Chounthirath

, Yang

, Friedenberg

, Smith

GA.

Children treated for lawn mower-related injuries in US emergency departments, 1990-2014. The American Journal of Emergency Medicine 2017;35(6), 893–898.

13.

Sekkay

, Imbeau

, Chinniah

, Dubé

, de Marcellis-Warin

, Beauregard

, Trépanier

Risk factors associated withself-reported musculoskeletal pain among short and long distanceindustrial gas delivery truck drivers. Appl Ergon (2018);72,69–87.

14.

García

, Sánchez

, Domingo

, Sebastián

MÁ.

Acost-benefit ergonomics methodology applied to service industry withdigital human modelling. European Journal of Industrial Engineering (2014)84, 533–553.

15.

Manghisi

, Uva

, Fiorentino

, Bevilacqua

, Trotta

, Monno

Real time RULA assessment using Kinect v2 sensor. Appl Ergon (2017);65, 481–491.

16.

Sen

, Sanjog

, Karmakar

A comprehensive review of work-related musculoskeletal disorders in the mining sector and scope for ergonomics design interventions. IISE Transactions on Occupational Ergonomics and Human Factors (2020)1–18.

17.

Grobelny

, Michalski

Preventing work-related musculoskeletal disorders in manufacturing by digital human modeling. International Journal of Environmental Research Public Health (2020);17(22), 8676.

18.

Besharati

, Daneshmandi

, Zareh

, Fakherpour

, Zoaktafi

Work-related musculoskeletal problems and associated factors among office workers. International Journal of Occupational Safety and Ergonomics: JOSE (2020);26(3), 632–638.

19.

Joshi

, Deshpande

Investigative study and sensitivity analysis of Rapid Entire Body Assessment (REBA). Int. J. Ind. Ergon (2020)79.

20.

Ghasemi

, Mahdavi

A new scoring system for the Rapid Entire Body Assessment (REBA) based on fuzzy sets and Bayesian networks. Int. J. Ind. Ergon (2020)80.

21.

Andreas

G-WJ

, Johanssons

Observational methods for assessing ergonomic risks for work-related musculoskeletal disorders. A scoping review. RevistaCiencias de la Salud (2018);16(SPE), 8–38.

22.

Diego-Mas

, Alcaide-Marzal

, Poveda-Bautista

Errors Using Observational Methods for Ergonomics Assessment in Real Practice. Human Factors (2017);59(8), 1173–1187.

23.

Sukadarin

, Deros

, Ghani

, MohdNawi

, Ismail

AR.

Postural assessment in pen-and-paper-based observational methods and their associated health effects: a review. International Journal of Occupational Safety and Ergonomics: JOSE (2016);22(3), 389–98.

24.

Berlin

, Neumann

, Theberge

, Örtengren

Avenues of entry:how industrial engineers and ergonomists access and influence humanfactors and ergonomics issues. European Journal of IndustrialEngineering (2014);8(3), 325–348.

25.

Pezzuolo

, Milani

, Zhu

, Guo

, Guercini

, Marinello

On-Barn Pig Weight Estimation Based on Body Measurements by Structure-from-Motion (SfM). Sensors (Basel, Switzerland). (2018);18(11) .

26.

Zhang

Microsoft kinect sensor and its effect. IEEE MultiMedia (2012);19(2), 4–10.

27.

Foreman

, Engsberg

JR.

The Validity and Reliability of the Microsoft Kinect for Measuring Trunk Compensation during Reaching. Sensors (Basel, Switzerland). (2020);20(24).

28.

Alazrai

, Momani

, Daoud

MI.

Fall detection for elderly from partially observed depth-map video sequences based on view-invariant human activity representation. Applied Sciences (2017);7(4), 316.

29.

Kim

, Choi

Performance comparison of user interface devices for controlling mining software in virtual reality environments. Applied Sciences (2019);9(13), 2584.

30.

Sung

N-J

, Ma

, Choi

Y-J

, Hong

Real-time augmented reality physics simulator for education. Applied Sciences (2019);9(19), 4019.

31.

, McGorry

, Chou

, Lin

, Chang

CC.

Accuracy of theMicrosoft Kinect for measuring gait parameters during treadmillwalking. Gait & Postu.(2015);42(2), 145–51.

32.

Siemens

Tecnomatix Jack and Process Simulate Human. In 2014.

33.

Siemens , Jack and Process Simulate Human Siemens Product Lifecycle Management Software. Zugriff am (2009);27, 2009.

34.

Zhang

, Wu

, Gao

, Chen

, Xv

Simulation and ergonomic evaluation of welders’ standing posture using Jack software. International Journal of Environmental Research Public Health (2019);16(22), 4354.

35.

, Tian

, Duffy

VG.

Performing ergonomics analyses through virtual interactive design: Validity and Reliability Assessment. (2012);22(3), 256–268.

36.

Martin

, Burkert

, Choi

, Wieczorek

, McGregor

, Herrmann

, Beling

PA.

In A real-time ergonomic monitoring system using the Microsoft Kinect, 2012 IEEE Systems and Information Engineering Design Symposium.(2012);201250–55.

37.

Diego-Mas

, Alcaide-Marzal

Using Kinect™ sensor in observational methods for assessing postures at work. Applied Ergonomics (2014);45(4), 976–985.

38.

Pfister

, West

, Bronner

, Noah

JA.

Comparative abilities of Microsoft Kinect and Vicon 3D motion capture for gait analysis. Journal of Medical Engineering & Technology (2014);38(5), 274–80.

39.

Dwyer

, Huckleby

, Kabbani

, Delano

, De Sutter

, Crawford

Ergonomic assessment of robotic general surgeons: a pilot study. Journal of Robotic Surgery (2020);14(3), 387–392.

40.

Haggag

, Hossny

, Nahavandi

, Creighton

In Real Time Ergonomic Assessment for Assembly Operations Using Kinect, UKSim 15th International Conference on Computer Modelling and Simulation.(2013);2013495–500.

41.

Cao

, Zhong

, Cao

, He

Physiological function assessment based on kinect V2. IEEE Access (2019);7, 105638–105651.

42.

Chiang

, Stephens

, Potvin

Retooling Jack’s Static Strength Prediction Tool. In SAE International 2006.

43.

Hovanec

Digital factory as a prerequisite for successful application in the area of ergonomics and human factor. Theoretical Issues in Ergonomics Science (2017);18(1), 35–45.

44.

Raschke

, Martin

, Chaffin

DB.

Distributed moment histogram: A neurophysiology based method of agonist and antagonist trunk muscle activity prediction. Journal of Biomechanics (1996);29(12), 1587–1596.

45.

Godwin

, Eger

, Corrigan

, Grenier

SG.

Classic JACK modelling of driver posture and line-of-sight for operators of lift-trucks. International Journal of Human Factors Modelling Simulation (2010);1(3), 259–270.

Ergonomics evaluation of lawn mower operator’s working posture using JACK software and kinect interface

Abstract

BACKGROUND:

OBJECTIVE:

METHOD:

RESULTS:

CONCLUSION:

Keywords

1 Introduction

2 Materials and methods

2.1 Rapid Upper Limb Assessment (RULA)

2.2 Static Strength Prediction (SSP)

2.3 Lower Back Analysis (LBA)

2.4 Development of human model using JACK

Table 1 Anthropometric characteristics of the operators Name Gender Age Standing height (cm) Body-weight (Kg) Operator-1 Male 21 166.40 63.5 Operator-2 Male 25 175.49 77.9 Operator-3 Male 30 176.00 79.3

3.1 RULA grand scores for each operator

3.2 Static Strength Prediction (SSP) for each operator

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Anthropometric characteristics of the operators

Name Gender Age Standing height (cm) Body-weight (Kg)

Operator-1 Male 21 166.40 63.5

Operator-2 Male 25 175.49 77.9

Operator-3 Male 30 176.00 79.3