Understanding the Musculoskeletal Demand of Ride-On Mowing Using Wearable Technology

Abstract

This study aimed to quantify the postures and muscle activity while parks and gardens workers operated ride-on mowers during a typical shift. Eight participants operated ride-on mowers in the same park but on different terrains (flat and undulating). Body postures and muscle activity were collected wirelessly and unobtrusively. Participants adopted a forward-flexed seated posture with the predominant movement being head rotation. Oscillatory movements (20–40° from neutral) of the thorax in all three planes of movement were noted in all participants. Low levels (<30% MVIC) of muscle activity were recorded in all muscles tested. These levels were elicited for most (>90%) of the recording time. Higher (>50% MVIC) activation levels were interspersed through the data, but these were not sustained. There was no difference in posture or muscle activity between the flat and undulating terrain. The forward-flexed posture combined with vibration can increase the risk of discomfort and injury in the low back while ride-on mowing. The low levels of muscle activity suggest participants did not actively brace for the occupational situation and task. The large inter-participant difference in posture attests to subjective variation to accommodate muscular stress, and this may not be optimal for injury mitigation.

1. Introduction

Ride-on mowers are often used in agricultural, landscaping and ground maintenance work, while the public also use such machines in domestic settings [1]. Research into the safe use of ride-on mowers has focused on roll over and amputation injuries as well as noise and vibration exposure [1,2]. A US epidemiological investigation showed nationally, more than 80,000 presentations to the emergency department were recorded yearly for the period between 2005 and 2015 [3]. These data do not clearly differentiate between push or ride-on mowers; however, the most common injuries reported were laceration (23.1%) followed by musculoskeletal strain and sprain injuries (18.8%) [3]. The risk of musculoskeletal injury in mowing has been examined via vibration analysis [4] but the kinematics and muscle activity while operating ride-on mowers has yet to be quantified. This leaves a gap in our knowledge of the total risk of using these machines.

The paucity of information regarding the musculoskeletal demands of ride-on mower use may, in part, be due to the challenges of performing ergonomic and biomechanical analyses. Traditionally, these forms of analyses would be performed in a laboratory setting or with outdoor, optical motion capture cameras. This setup limits our ability to perform the analyses with environmental fidelity (when in a laboratory) or with only limited capture volumes (when using outdoor cameras).

Wireless inertial measurement units (IMUs) and surface electromyography (sEMG) technology provides the opportunity to assess kinematics and muscle activity unobtrusively, and in a work-specific context. IMUs contain a combination of three-dimensional accelerometers, rate gyroscopes, inclinometers and a magnetometer. They can be used singularly to assess the orientation of an object they are affixed to or as an array of interconnected sensors where the measurement of human segment kinematics is enabled. In terms of their validity in biomechanical studies, spinal kinematics from IMUs have been shown to be accurate compared to traditional optoelectronic motion capture [5,6]. sEMG has been commonly used to assess muscle activity in human movement. The wireless advancement of these sensors has allowed the application of this technology to a variety of more previously challenging environments such as in firefighting, during rapid sporting movements, and in long-duration analysis such as office desk use. Recently, in combination, the suite of IMUs and sEMG has been used to quantify the posture and muscle activity of waste drivers [7], giving invaluable insight into a challenging job task while also not interfering with the natural process of the work.

Considering the gaps in our knowledge with ride-on mowing, especially in the area of musculoskeletal demand, this project aimed to quantify the specific spinal and upper body postures as well as selected muscle activity when participants operated ride-on mowers during a typical mowing session (30 min). This information has the potential to increase the understanding of the work demand of ride-on mowing, allowing interventions to mitigate accident and injury to be formulated.

2. Materials and Methods

2.1. Participants

Eight participants (2 female; 6 male; mean ± SD; age: 35 ± 8 years; mass: 82 ± 12 kg; height: 1.73 ± 0.1 m; experience: 3 ± 2.3 years) volunteered for this study. Participants were recruited from a pool of parks and gardens workers within a local government organisation and were included if they presented with no current or recent (within six months) musculoskeletal injury and were trained in the operation of a ride-on mower. Participants in the study gave informed consent prior to the commencement of any data collection. Ethical approval (RDHS-62-15) was obtained from the Institutional Human Research Ethics Committee prior to all participant recruitment and testing procedures.

2.2. Data Collection

An initial observation of the task revealed marked differences in the types of terrains mowed. Thus, it was decided to conduct the analysis in a single location (Remembrance Park, Banksia & Studley Streets, Heidelberg, Victoria, Australia) where both flat and hilly terrains are available to be mowed. A cross-sectional study design was used to examine three-dimensional spinal and shoulder postures and muscle activity from three muscles while participants operated the ride-on mower. Participants performed two types of mows, a flat terrain mow and an undulating terrain mow in the same park. All participants wore their typical work wear which comprising self-supplied overalls, safety boots, and eye and ear protection. They also used their council-supplied ride-on mower (Kubota F3690SN-72, Osaka, Japan). No familiarisation trial was conducted as participants were well versed with their equipment and the task.

Wireless IMUs (Noraxon Myomotion, Noraxon, Scottsdale, AZ, USA) was used to analyse the postures and joint angles associated with operating the mowers. An array of six wireless and lightweight sensors were attached to the participants’ head (on the occiput in line with the midline of the body), over C7, over T12, on the left and right upper arms (lateral aspect) and on the pelvis (over L5/S1) according to manufacturer specifications and using manufacturer-supplied straps with pouches or double-sided tape. All kinematic data were collected at 100 Hz.

From the array of IMUs, a selection of joint angles was extracted using manufacturer-supplied software (MR 3.10, Noraxon, Scottsdale, AZ, USA). This included flexion/extension, lateral flexion, and axial rotation angles of the lumbar, thoracic and cervical (head and neck) spine. These data were then processed to obtain an average posture adopted during the data collection period. Further, a range of movement (ROM) value for all the joint angles was obtained by calculating the absolute range of the joint from the peak excursion adopted during the capture period.

Muscle activation from three specific muscles was measured using wireless surface electromyography (sEMG) (Noraxon DTS, Noraxon, Scottsdale, AZ, USA) at 1500 Hz. The muscles of interest were m. flexor carpi radialis, m. upper trapezius and m. lumbar erector spinae. All sEMG readings were measured unilaterally from the right side of the body as all participants reported they were right-handed. The locations were found using industry-standard guidelines to ensure accurate and consistent placement [8] (Figure 1).

Each muscle site was prepared to ensure a clean and accurate sEMG signal could be recorded. This involved the area being shaved, gently abraded and wiped with a sterile swab. Once all sEMG equipment was affixed and tested, each participant completed a maximal voluntary isometric contraction (MVIC) of each of the muscles being recorded. This process was completed by pulling on an anchored, inextensible exercise strap (TRX, Fitness Anywhere LLC, San Francisco, CA, USA) in specific reference movements according to published guidelines [8].

All sEMG signals were full-wave rectified and root-mean squared smoothed using a 200 msec window [7]. A peak value was obtained from the MVIC trials, and these were used to normalise the trials for each participant. To further understand the muscle activity, average and peak normalised sEMG were calculated using manufacturer-supplied software (MR 3.10, Noraxon, Scottsdale, AZ, USA). Also, muscle activity data were segregated into the time spent in a specific zone. For this study, we performed a secondary time-in-zone analysis using the amplitude distribution profile of <30% MVIC, 30–50% MVIC and >50% MVIC. These data showed how much time participants were spending in each activation zone. sEMG processing procedures were adopted from previous work investigating muscle activity while driving waste vehicles [7].

2.3. Data Analysis and Statistics

All data were segregated into the terrain mowed where group averages and standard deviations were obtained and reported. As the sEMG data appeared quite varied, we also calculated interquartile ranges for these. To answer the question of whether terrain caused a change in posture or muscle activity, we used a one-way ANOVA to contrast the average postures, ROM, average and peak normalised sEMG and time in specific sEMG zones (Microsoft Excel 2013, Microsoft, Redmond, WA, USA). Statistical significance was set at p ≤ 0.05.

3. Results

The average posture analysis revealed participants adopted forward flexion posture originating at the lumbar spine (45° lumbar flexion; Figure 2). The range of motion analysis showed the largest range of movement was cervical axial rotation with participants twisting their head over 70° to both the left and right during the data collection period (Figure 3). Much smaller (22–42°) ranges were recorded in the other spinal segments and planes of motion (

Table 1. Average (±SD) and range (±SD) of spinal postures adopted while ride-on mowing.

	Cervical			Thoracic			Lumbar
	Flexion	Lateral Flexion	Axial Rotation	Flexion	Lateral Flexion	Axial Rotation	Flexion	Lateral Flexion	Axial Rotation
Average Posture (°)	−5.6 (±8.2)	2.5 (±13.9)	−1.3 (±44.3)	−6.7 (±16.5)	3.1 (±14)	10.8 (±32.9)	45.6 (±6.2)	5.5 (±6.7)	−2.2 (±19.5)
Range (°)	41.7 (±9.3)	41.1 (±8.1)	141.8 (±13.3)	22.3 (±9.3)	18.8 (±11.3)	35.8 (±9.2)	31.3 (±13.2)	38.1 (±6.2)	34.6 (±11.3)

). The ANOVA revealed no significant differences in average postures or ROM when these comparisons were made between the flat and hilly terrains.

In our trials, m. flexor carpi radialis displayed the highest average (14%MVIC in hilly) and peak muscle activity (98%MVIC in hilly), while m. upper trapezius displayed the lowest average (5% in hilly) and peak (33% in flat) muscle activity (

Table 2. Normalised sEMG for each of the muscles investigated.

	Average ± SD (%MVIC)		Interquartile Range (%MVIC)		Peak ± SD (%MVIC)		Interquartile Range (%MVIC)
	Flat	Hilly	Flat	Hilly	Flat	Hilly	Flat	Hilly
Flexor carpi radialis	12 ± 7	14 ± 7	8	4	81 ± 38	98 ± 47	64	40
Upper trapezius	6 ± 3	5 ± 3	4	4	33 ± 22	41 ± 19	11	20
Lumbar erector spinae	12 ± 7	14 ± 4	9	4	65 ± 52	66 ± 15	28	15

). There were instances of higher (>50% MVIC) activation recorded in all muscles, but these were not sustained for any length of time. Specifically, most of the muscle activity recorded was <30% MVIC. The time-in-zone analysis revealed that this low level of muscle activity accounted for the majority (>90%) of the assessment time (

Table 3. Percentage of trial time spent in the different muscle activation zones.

	Flexor Carpi Radialis	Upper Trapezius	Lumbar Erector Spinae
<30% MVIC	92 ± 4.5%	99 ± 1.3%	98 ± 3.8%
30–50% MVIC	5 ± 3.6%	0.5 ± 1.1%	1 ± 2.1%
>50% MVIC	3 ± 1.8%	0.5 ± 0.1%	1 ± 1.8%

). The ANOVA showed no significant differences in mean and peak muscle activation levels between flat and hilly terrains.

4. Discussion

The main findings of our study showed that while ride-on mowing, our participants adopted a forward-flexed posture originating at the lumbar spine. They also used large ranges of movement to rotate their heads while mowing. The sEMG readings showed that low-level muscle activity interspersed with short bursts of higher activation were recorded from the trunk and upper limb muscles.

The forward-flexed posture of the trunk recorded in this study is similar to other reports of driving heavy vehicles [7,9]. These postures in isolation are not associated with the risk of developing musculoskeletal injury. However, when factors such as whole-body vibrations, shocks and jolts that come with mowing uneven, bumpy terrain, the risk of musculoskeletal injury may increase [10]. A study of grass harvesting using a tractor showed significant increases in vibration especially where working characteristics were severe [11]. The risk of developing musculoskeletal injury in these situations is further increased if the seat posture and exposure to vibration/undulations are prolonged (>2 h) [12]. Johnson and colleagues [13] showed active suspension seats significantly decreased vibration in heavy trucks compared to air-suspended seats.

The predominant movement recorded in our study was axial rotation of the head about the trunk. Similar kinematics have been recorded in drivers of waste collection vehicles [7] and also in bus drivers [14]. These studies have attributed this movement to the attention demand of the work, and this is also similar in operating ride-on mowers. The mowers used in our study did not have any means of rearward vision. Thus, operators had to rotate their head, sometimes to near end-of-range, to facilitate visual information.

Data from the ride-on mowers suggest participants may have encountered a combination of vibration of the mower as well as jolts from the bumps in the terrain being mowed. The ranges of movement (>30°) in all three planes in the lumbar spine (Table 1) along with the low levels of muscle activity (Table 2 and Table 3) recorded suggest participants did not brace for these perturbations [15]. Bracing for sudden perturbations has been shown to reduce intersegmental movement in the lumbar spine [15]. Education and specific practice about bracing may be valuable when operating ride-on mowers, and managers of this workforce may wish to schedule targeted training in this area.

The vibration from the mower and the sudden perturbations from the terrain combined with the pedestal-like nature of the ride-on mower seat may have compromised participants’ ability to stabilise their posture. Slota and colleagues [16] showed impaired trunk postural control after whole-body vibration in unstable sitting. Also, given the open nature of the ride-on mower seat, participants would have no opportunity to brace themselves. This can be evidenced from the low levels of muscle activity exhibited while mowing. Manufacturers of ride-on mowers may wish to lower the seat of the mower and better design seats to allow participants to brace themselves.

There was a large inter-participant difference in seated postures, and this can also be attributed to the vibration and undulation of the terrain and mower [17]. More specifically, the oscillatory movement in the spinal kinematics measured during mowing may be attributed to these factors.

This study used wearable sensors that allowed field-based kinematic and electromyographic analysis in a task that would have been impossible to simulate in a laboratory. Our study was limited by the small and homogenous nature of our sample. We also did not consider individual participant characteristics such as age, experience, sex and body mass index in our analysis. Further, the design of our study does not allow causative factors of injury to be assessed. Future investigations may wish to consider these limitations and study this important workforce in a longitudinal manner such that factors associated with increase injury risk can be ascertained.

Further research in ride-on mowers could examine whole-body vibration or attempt to manipulate standard fixtures on the mower and test the resulting changes in musculoskeletal load. Also, future designs of ride-on mowers may wish to incorporate mirrors or rearward facing cameras to enhance operator visual information and minimise the frequency of cervical rotations.

5. Conclusions

Our study shows operators of ride-on mowers are faced with several stressors that may impact their musculoskeletal system. Solutions such as active damping seats, rear vision, targeted training strategies and appropriately scheduled breaks can have a positive impact on the health and well-being of this vital work group. Our findings were limited by a small and homogenous sample, and the design of the study did not allow injury causation to be examined. Future research may wish to focus on these as well as examine the role whole-body vibration has on injury in this vital cohort of workers.