PATH法と階層的タスク分析による手作業ライチ収穫時の人間工学的リスク評価：観察横断研究
2026 年 48 巻 2 号 p. 71-85
詳細
2026 年 48 巻 2 号 p. 71-85
Musculoskeletal disorders (MSDs) are prevalent among agricultural workers, particularly during harvesting activities involving awkward postures and heavy manual handling. Despite Thailand’s significant role in global lychee production, ergonomic risks in lychee harvesting remain under-investigated, even though exposure is evident. This cross-sectional observational study was conducted during the 2022 lychee harvest season across eight orchards in Phayao Province, Thailand. A total of 104 male harvesters were systematically video-recorded during full-task execution. Hierarchical Task Analysis (HTA) was used to segment workflows into goal-directed sub-tasks, and ergonomic exposures were assessed using the Posture, Activity, Tools, and Handling (PATH) method, with postures coded as neutral or non-neutral using a structured checklist. The harvesters were predominantly older working-age adults (mean age 53.5 years) who worked long hours (7.65 hours/day), often carrying 9.5 kg baskets unilaterally on one shoulder. Postural analysis revealed frequent ergonomic risks: 59% of neck postures and 58% of trunk postures were non-neutral, including 15% with severe forward flexion and 24% involving lateral bending or twisting. Arm elevation above shoulder height was common during climbing and fruit detachment. Spatial layout imposed long travel distances (over 30 meters), contributing to cumulative fatigue. Ladder height, tool design, and task repetition further exacerbated biomechanical strain. Lychee harvesting imposes significant ergonomic risks due to non-neutral postures, load asymmetry, and task repetition. Evidence supports the urgent implementation of ergonomic interventions, including bilateral-carry basket redesign, improved ladder safety, and participatory workflow reorganization to reduce WMSD risks in seasonal fruit agriculture.
Musculoskeletal disorders (MSDs) are a global health concern, particularly among agricultural workers who perform physically demanding tasks under challenging environmental conditions [1]. Principal occupational risk factors include repetitive movements, awkward postures, and manual handling [1, 2]. The burden of work-related MSDs (WMSDs) is especially high in countries such as Thailand, where agriculture is a major source of income [3]. Prevalence in Southeast Asia ranges from 78.31% to 88.39% [4], with a Thai systematic review reporting a rate of 67.8% [5].
MSDs are particularly common during harvesting, when tasks involving lifting, bending, and reaching overhead intensify. Thai farmers report more MSDs in the lower back, knees, shoulders, and wrists during harvesting than planting [6]. Working posture and height influence severity: reaching overhead contributes to neck and shoulder pain; while squatting and bending are linked to discomfort in the shoulders and back [7]. Psychosocial factors such as long hours and time pressure also elevate risk. Low back pain (LBP) is the most frequently reported MSD, followed by pain in the knees, shoulders, and neck [8]. However, most studies rely on self-reports and lack objective, field-based assessment [9].
Lychee harvesting in northern Thailand presents unique ergonomic risks. The Asia-Pacific region produces over 95% of global lychee output, with northern Thailand as a major contributor [10]. Despite research on WMSDs among Thai farmers, most studies have focused on rice and other crops, leaving lychee harvesting under-examined [3–6].
Recent evidence highlights the severity of the issue. Among ethnic lychee and longan harvesters, 99.5% reported MSD symptoms in at least one body region, especially in the hands, shoulders, and neck, due to prolonged static postures, repetitive motion, and heavy lifting [11]. A study in Phayao Province found that 70.5% of lychee farmers performed high-risk tasks such as ladder climbing and prolonged overhead reaching, with MSDs concentrated in the lower back, upper back, and dominant wrist [12]. These findings emphasize the need for ergonomic assessments and targeted interventions.
Such ergonomic assessments can be guided by Hierarchical Task Analysis (HTA), which provides a structured framework for decomposing complex harvesting workflows into discrete, goal-directed sub-tasks. HTA is widely used in ergonomics to analyze manual tasks, enhance risk identification, and improve workflow efficiency [13]. Its recent applications in agricultural and forestry contexts underscore its relevance in field-based ergonomic research [14]. In this study, HTA served as the conceptual foundation for mapping harvest activities and informing subsequent ergonomic evaluation.
The Posture, Activity, Tools, and Handling (PATH) method is a structured observational approach for quantifying ergonomic risks across various industries, including construction, demolition, fishing, and agriculture [15–18]. PATH captures non-neutral postures, load handling, and repetitive tasks associated with MSD risks. For example, construction workers spend up to 96% of their time with elevated elbows [17], and apple harvesters carry heavy loads during more than 75% of their work time [18]. The method has been refined in horticulture to include interval-scaled posture coding [19, 20]. In agricultural settings, PATH effectively identifies MSD risks in repetitive, overhead, and load-bearing tasks [16, 20, 21]. These insights inform strategies to reduce physical strain, especially for ladder use and prolonged static postures [16, 21]. Despite its extensive application across various agricultural sectors, the use of PATH in lychee harvesting remains undocumented.
This study addresses that gap by applying the PATH framework, guided by HTA, to evaluate ergonomic risks among lychee harvesters in northern Thailand. By integrating posture, task duration, tool use, and handling techniques, this approach identifies modifiable risk factors and offers an evidence-based foundation for improving worker safety in this under-researched agricultural sector.
Study Design and Setting
This cross-sectional observational study was conducted during the 2022 lychee harvest season, spanning March to June. Field observations were carried out in eight commercial orchards located across three sub-districts in Mae Chai District, Phayao Province, which represents one of the largest lychee cultivation zones in Northern Thailand (Figure 1). The study protocol was reviewed and approved by the University of Phayao Human Ethics Committee (Approval No. UP-HEC-1.2/022/64). All participants provided written informed consent prior to enrollment in the study.
Participants
A total of 104 male lychee harvesters were recruited using purposive-convenience sampling, with assistance from local health volunteers. The participants were pre-scheduled for video recording to ensure full-cycle observation, from task initiation to completion. Inclusion criteria were: (1) age ≥ 20 years; (2) at least one year of orchard work experience; (3) currently active in lychee harvesting; and (4) directly responsible for performing both harvesting (e.g., climbing and fruit detachment) and transporting (e.g., carrying baskets) tasks during the study period. Those performing only sorting or loading, or with acute injuries or physical limitations, were excluded.
Video Recording and Observer Training
Work cycles were recorded using a digital camcorder at 25 frames per second. Cameras were positioned laterally at a distance of approximately 3 to 5 meters from the worker to obtain a clear, unobstructed side view of trunk and limb postures. Environmental lighting and recording angles were adjusted prior to each session to ensure adequate visual clarity for postural assessment.
Two ergonomists underwent calibration using sample video clips based on the adapted PATH checklist. A random subsample of 20 clips (approximately 19% of the total), stratified by sub-task, was independently coded. Discrepancies in classification were resolved through discussion to ensure coding consistency.
Hierarchical Task Analysis
Based on prior research in the same geographic region [12], the lychee harvesting process was categorized into four stages: (1) harvesting, performed either by climbing ladders or standing on the ground; (2) transporting, which involved carrying baskets to sorting stations; (3) sorting, including grading and packing; and (4) loading, which consisted of placing packed boxes onto transport vehicles. While all of the stages were observed, only harvesting and transporting were selected for ergonomic assessment. These two stages were chosen because they were performed by all of the participants, involved the greatest physical demands, and provided video data of sufficient duration and consistency for posture coding. Sorting was excluded as it was performed by a different group of workers not included in the analytic sample. In contrast, loading was excluded because it consisted mainly of occasional lifting of packed boxes, with shorter duration and lower repetition compared to harvesting and transporting.
A hierarchical task analysis (HTA) was developed post hoc, based exclusively on video observations of the participants engaged in harvesting and transporting. Instead of following a predefined structure, task segmentation was derived inductively from observed patterns, consistent with HTA principles that emphasize goal-directed sub-units and logical sequencing [13]. Two ergonomists, each with over five years of field experience, independently reviewed the footage to identify repeating elements, decision points, and task loops. Through consensus, they created a 12-step HTA chart (Steps 1.1 to 2.4), which covered actions such as ladder setup, climbing, fruit detachment, basket handling, and load transport. This structured breakdown, as illustrated in Figure 2, informed the organization and interpretation of the subsequent ergonomic analysis.
PATH Assessment Procedures
The Posture, Activity, Tools, and Handling (PATH) method was applied to evaluate ergonomic exposures within sub-tasks identified through the HTA. Each sub-task was mapped to its dominant PATH domain (e.g., posture, activity, handling), as outlined in Table 1.
| HTA Step | PATH Dimension | Manifestation in Task |
|---|---|---|
| 1.1 Prepare & Position Ladder | Posture, Handling | Awkward trunk/arm posture when lifting, carrying and setting ladder into place |
| 1.2 Attach Basket Hook | Posture, Handling | Reaching and securing basket hook at waist or on ladder rail under load |
| 1.3 Climbing up/down | Posture, Activity | Repetitive trunk flexion/extension and arm movements as worker ascends and descends |
| 1.4 Breaking lychee bunches | Posture, Tools | Sustained wrist flexion and grip force when twisting stems or operating hand shears |
| 1.5 Dropping into basket | Posture, Activity | Frequent arm elevation and lowering motions combined with hand–eye coordination |
| 1.6 Basket-Full Check | Posture | Brief stooping or leaning posture to visually assess basket fullness |
| 1.7 Move to Next Tree | Posture, Handling | Trunk bending and load carriage while relocating ladder and basket to the next tree |
| 1.8 Dismount & Pack Away Ladder | Posture, Handling | Stooping and lift-carry of ladder when removing hook and storing safely |
| 2.1 Inspect Basket Load | Posture, Handling | Stooping to verify load distribution and balance of basket |
| 2.2 Lift Basket | Posture, Handling | Squat-to-stand trunk posture under load during initial lift of filled basket |
| 2.3 Carry to Sorting Station | Posture, Handling | Sustained upright posture under load over walking distance to sorting area |
| 2.4 Place Basket Down | Posture, Handling | Controlled squat/stoop and lowering action to set basket onto table or ground |
HTA: Hierarchical Task Analysis, PATH: Posture, Activity, Tools, and Handling method.
Two levels of analysis were performed. First, a detailed sub-task–specific analysis targeted Steps 1.3 (Climbing), 1.4 (Breaking Lychee Bunches), and 1.5 (Dropping into Basket), each averaging ≥1 minute per cycle. These were continuously coded using a 1-second interval, pilot-tested to balance feasibility with resolution, to capture rapid trunk and upper limb posture changes. These three sub-tasks were selected because they occurred consistently across all of the harvesting cycles and participants, and represented the primary ergonomic risks of fruit harvesting, including trunk flexion, overhead reaching, and repetitive load transfer. Other sub-tasks were shorter or irregular, providing insufficient duration for detailed posture coding.
To quantify task repetition, each complete harvest cycle was reviewed to count the frequency of observable actions within these sub-tasks, such as the number of ladder ascents or descents, bunch-breaking movements, and dropping actions per basket load. Repetition counts were extracted directly from full-cycle video recordings and used to characterize the degree of repetitive motion during harvesting.
Second, a broader analysis covered the distribution of postures across all of the harvesting tasks. The observed postures were classified as neutral or non-neutral using a dichotomous checklist adapted from Fulmer et al (2002) [21], following the original PATH definitions [22]. Thresholds such as trunk flexion >45°, arm elevation above shoulder height, and knee flexion >35° were applied as visual guidelines for categorical coding rather than precise angles.
Additional PATH domains were evaluated to complement postural data. Task durations were recorded for representative actions (e.g., basket handling, ladder holding), tool characteristics such as basket and ladder dimensions were documented, and load weights were measured directly for handling tasks. These assessments enabled a more comprehensive profile of the ergonomic demands.
Although PATH was originally designed for work sampling in non-cyclic tasks, it has since been refined and widely applied in agricultural and other non-routinized occupations. In this study, it was adapted for structured, video-based observation of harvest cycles to improve temporal accuracy while preserving its focus on task-level exposures. PATH was selected over alternative posture assessment methods because it extends and refines the OWAS framework by systematically linking postures to activities and incorporating additional posture codes, such as for the neck and varied trunk flexion levels [15, 18, 22]. Unlike rapid upper limb assessment (RULA), which relies on single-time posture snapshots and is more suited for static or short-cycle work, PATH employs a work-sampling approach to capture whole-task dynamics, making it particularly suitable for evaluating ergonomic risks in lychee harvesting [16, 20, 21].
Data Processing and Reliability
Observational data were entered in Microsoft Excel and analyzed in SPSS version 28. The main outcome was the percentage of time spent in each posture category during physically intensive sub-tasks identified via HTA (1.3, 1.4, 1.5), focusing on trunk and upper limbs. This was calculated by dividing non-neutral posture frames by total frames per task and segment. Participant-based variables (e.g., demographics, work history) were recorded for n=104 harvesters. Task-based variables (e.g., ladder usage frequency, distances between trees) generated a greater number of observations (up to 135 cycles), since some participants performed more than one harvest within the observation period.
Inter-rater reliability was assessed using Cohen’s kappa (κ) on a stratified sample of 20 harvest cycles, yielding κ = 0.82 (substantial agreement). A pilot test of 15 cycles prior to full data collection produced a Cronbach’s alpha of 0.80, indicating acceptable checklist consistency.
Statistical Analysis
Descriptive statistics were used to summarize all of the relevant variables, including demographic characteristics, work patterns, ergonomic equipment dimensions, task durations, physical load weights, and repetition counts. For postural analysis, the proportion of time spent in each posture category (e.g., trunk neutral, arm elevation) was computed and reported by task type and body region. Results are presented as means, standard deviations, ranges, and percentages.
Participant Characteristics
The study included 104 participants with a mean age of 53.50 ± 8.78 years, weight of 62.09 ± 8.20 kilograms, height of 163.48 ± 4.67 centimeters, and BMI of 23.24 ± 3.03 kg/m². The participants had a mean of 17.54 ± 7.56 years of experience in lychee harvesting, working 14.12 ± 5.02 days per month and 7.65 ± 1.69 hours per day, with 4.07 ± 1.31 minutes spent per harvest cycle (Table 2).
| Variable | Mean ± SD or Percentage (%) |
|---|---|
| Age (Mean ± SD) | 53.50 ± 8.78 years |
| Weight (Mean ± SD) | 62.09 ± 8.20 kilograms |
| Height (Mean ± SD) | 163.48 ± 4.67 centimeters |
| Body Mass Index (BMI, Mean ± SD) | 23.24 ± 3.03 kg/meters2 |
| Work Experience in Lychee Harvesting (Mean ± SD) | 17.54 ± 7.56 years |
| Working Days per Month (Mean ± SD) | 14.12 ± 5.02 days |
| Working Hours per Day (Mean ± SD) | 7.65 ± 1.69 hours |
| Time per Harvest Cycle (Mean ± SD, minutes) | 4.07 ± 1.31 minutes |
| Usage Frequency of Ladder Crutches (n=135) | 121, 89.63 % |
| Preferred Carrying Side for Loaded Baskets (n=104) | |
| - Right side shoulder | 54, 51.92 % |
| - Left side shoulder | 32, 30.77 % |
| - Front of body | 18, 17.31 % |
| Ladder Usage per Harvest Cycle (n=104) | |
| - Once | 73, 70.19 % |
| - Twice | 22, 21.15 % |
| - Thrice | 9, 8.65 % |
n=104 refers to the number of participants. n=135 represents repeated harvest cycles, as some participants performed more than one harvest during the observation period.
Task Characteristics and Postures
Basket handling showed that 51.92% of the participants carried loaded baskets on the right shoulder, 30.77% on the left shoulder, and 17.31% in front of the body. Ladder use was common: 70.19% climbed once per harvest cycle, 21.15% climbed twice, and 8.65% climbed three times, with 89.63% using crutches to stabilize the ladders (Table 2, Figure 3).
Representative postures during harvesting included balancing ladders, climbing with parallel trunk alignment, overhead or oblique reaching for lychee clusters, and bending or lateral reaching when placing fruit into baskets. Postural adjustments varied with basket height and fullness (Figure 4). Distances and heights are summarized in Table 3: the mean distance from the collection point to the first tree was 32.27 m (range: 12.50–45.10), the return distance from the last tree was mean 27.53 m (10.12–45.10), and the distance between successive trees was 8.12 m (6.18–10.26). The mean harvesting height at shoulder level was 3.48 m (2.40–4.71).
| Distance/Height | Mean (m) | Range (m) |
|---|---|---|
| Distance from Collection Point to First Lychee Tree (n=104) | 32.27 | 12.50–45.10 |
| Distance Between Successive Plants (n=31) | 8.12 | 6.18–10.26 |
| Distance from Last Tree to Collection Point (n=104) | 27.53 | 10.12–45.10 |
| Height of Harvested Points (Shoulder Level, n=135) | 3.48 | 2.40–4.71 |
n=104 refers to participants. n=31 represents additional distances observed when participants harvested more than one successive tree (total 135 observations). Range values represent minimum–maximum distances (m).
Detailed analysis of three prolonged sub-tasks (Steps 1.3–1.5) showed that trunk postures were neutral in 97% of climbing tasks but more varied during breaking and dropping, with mild flexion in 25% and 11% of cases, and severe flexion in 21% and 24%. Lateral bending or twisting was observed in 24% of breaking tasks and 21% of dropping tasks. Arm postures were mostly neutral during dropping (68%), while raising one arm was recorded in 34% of climbing and 42% of breaking tasks; both arms raised occurred in 9% of climbing and 27% of breaking tasks but not in dropping. In addition, repetition frequencies showed that each harvester performed a mean of one to two ladder ascents or descents, twelve bunch-breaking movements, and ten dropping actions per harvest cycle (Table 4).
| HTA Step | Trunk Neutral (%) | Trunk Non-Neutral (%) | Arm Neutral (%) | One-Arm Raised (%) | Two-Arms Raised (%) | Repetition per cycle (times; mean ± SD [range]) |
|---|---|---|---|---|---|---|
| 1.3 Climbing up/down | 97 | 3 | 57 | 34 | 9 | 1.3 ± 0.6 [1–3] † |
| 1.4 Breaking lychee bunches | 14 | 86 | 31 | 42 | 27 | 12.3 ± 3.5 [6–20] ‡ |
| 1.5 Dropping into basket | 25 | 75 | 68 | 32 | 0 | 9.8 ± 3.2 [4–17] ‡ |
Steps 1.3–1.5 were selected because they were prolonged and physically demanding sub-tasks, generating consistent data for second-by-second posture coding.
† Climbing repetition per cycle reflects ascent counts (once, twice, or thrice); median and range are shown due to skewed distribution, with proportions detailed in Table 2 (70.2% once; 21.2% twice; 8.7% thrice).
‡ Repetition counts for all three sub-tasks were obtained from full-cycle video recordings, with values expressed as mean ± SD and range of repetitions per harvest cycle.
Aggregated distributions across all of the harvesting tasks are presented in Figure 5. Trunk postures were neutral in 42% of observations, while 58% were non-neutral, including moderate and severe flexion, lateral bending/twisting, and combined movements. Neck postures were non-neutral in 59% of cases, compared with 41% neutral. Legs were predominantly neutral (68%), with non-neutral categories such as bent legs, squatting, or one leg lifted occurring less frequently. Arm postures were neutral in 53%, followed by one arm raised and both arms raised. The definitions of categories A–E used in the figure are provided in the Figure 5 legend.
The X-axis represents posture categories (A–E) and the Y-axis the proportion of observations.
Posture categories are labeled according to PATH coding:
•Trunk: A = neutral (<20° flexion, minimal lateral bending/twisting), B = moderate flexion (20–45°), C = severe flexion (>45°), D = lateral bending/twisting (>20°), E = combined flexion and twisting.
•Neck: A = neutral (<30° flexion/lateral bending or <45° twisting), B = non-neutral (>30° flexion/lateral bending or >45° twisting).
•Legs: A = neutral (knee flexion <35°), B = one leg in the air, C = bent legs (knee flexion ≥35°), D = squatting.
•Arms: A = both elbows below shoulder height, B = one arm raised above shoulder height, C = both arms raised above shoulder height.
PATH: Posture, Activity, Tools, and Handling.
Tool Dimensions and Task Demands
Harvesting tools varied across orchards (Table 5). Baskets had a mean top diameter of 51.67 cm (40.00–73.00 cm), bottom diameter of 26.73 cm (25.00–30.00 cm), and depth of 40.47 cm (35.00–85.00 cm). Rope and hook heights measured a mean of 41.53 cm (40.00–46.00 cm) and 19.00 cm (16.00–31.00 cm), respectively. Ladders had a mean height of 5.88 m to the last rung (4.02–6.50 m) with 11.05 rungs (7.00–12.00), while the width and length of slots had mean values of 0.32 m (0.25–0.50 m) and 0.43 m (0.12–0.54 m), respectively. A mean of 6.94 rungs were used during climbing. Crutches measured 4.21 m (4.00–5.00 m).
| Component | Mean | Range | |
|---|---|---|---|
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/48_71_6.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
|
Basket | ||
| Top Diameter (cm) | 51.67 | 40.00–73.00 | |
| Bottom Diameter (cm) | 26.73 | 25.00–30.00 | |
| Depth (cm) | 40.47 | 35.00–85.00 | |
| Rope Height (cm) | 41.53 | 40.00–46.00 | |
| Hook Height (cm) | 19.00 | 16.00–31.00 | |
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/48_71_7.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
|
Ladder | ||
| Height to Last Rung (m) | 5.88 | 4.02–6.50 | |
| Number of Rungs | 11.05 | 7.00–12.00 | |
| Width of Slot (m) | 0.32 | 0.25–0.50 | |
| Length of Slot (m) | 0.43 | 0.12–0.54 | |
| Rungs at Climbing | 6.94 | 5.00–9.00 | |
| Length of Crutches (m) | 4.21 | 4.00–5.00 |
Range represents minimum–maximum values.
Task durations and load weights are presented in Table 6. Unloading baskets required 0.85 minutes (0.37–1.58), with a mean load of 1.57 kg (1.20–1.80). Loading baskets required 0.98 minutes (0.42–1.83), with a mean weight of 9.56 kg (5.40–12.00). Holding ladders lasted 1.12 minutes (0.53–2.38), with a mean load of 5.39 kg (4.10–6.20), while holding crutches lasted 1.08 minutes (0.32–1.87), with a mean load of 2.54 kg (1.80–3.00).
| Parameter | Duration (Minute) | Weight (kg) | ||
|---|---|---|---|---|
| Mean | Range | Mean | Range | |
| Holding the Basket | ||||
| - Unloading the Basket | 0.85 | 0.37–1.58 | 1.57 | 1.20–1.80 |
| - Loading the Basket | 0.98 | 0.42–1.83 | 9.56 | 5.40–12.00 |
| Holding the Ladder | 1.12 | 0.53–2.38 | 5.39 | 4.10–6.20 |
| Holding the Crutches of the Ladder | 1.08 | 0.32–1.87 | 2.54 | 1.80–3.00 |
Range represents minimum–maximum values.
Participant Characteristics and Workload Demands
The participants in this study were predominantly older working-age adults and had extensive lychee-harvesting experience. Despite their expertise, they performed repetitive harvesting cycles for extended hours each day, contributing to significant cumulative biomechanical load. These findings align with regional trends showing that age and employment duration are major risk factors for work-related musculoskeletal disorders (WMSDs), especially in tasks involving repetition and heavy lifting. A systematic review in Southeast Asia reported that workers over 40 years old with more than ten years of experience are particularly vulnerable to WMSDs [4]. Similarly, a recent meta-analysis from low- and middle-income countries found that long working hours and prolonged employment were significantly associated with higher MSD prevalence, with repetitive movements and heavy lifting identified as primary occupational exposures linked to musculoskeletal pain [23].
Postural Risks During Harvesting Tasks
The harvesting tasks frequently involved non-neutral postures, with notable instances of severe trunk flexion, twisting, and lateral bending. The workers also regularly elevated their arms above shoulder height during climbing and fruit detachment, reflecting the overhead demands typical of tree-fruit harvesting. Similar postural patterns have been observed in oil palm work, where a majority of harvesters engage in overhead tasks [24], and in apple picking, where approximately two-thirds of work time is spent in awkward positions such as stooping and reaching [25].
Overhead work has been consistently associated with increased MSD severity in the neck, shoulders, upper back, and elbows [7]. In our study, 59% of the neck postures were non-neutral, underscoring the ergonomic stress placed on cervical structures. This is supported by biomechanical findings showing elevated upper trapezius activity during neck extension, particularly in tasks requiring overhead reaching [7, 26].
These results align with ergonomic evaluations across fruit-picking sectors, where prolonged overhead reaching, trunk flexion, and ladder handling dominate daily activity and contribute to elevated MSD risk [7, 27]. In blueberry harvesting, over 60% of postures were classified in the moderate-to-high risk category, especially for shoulders and trunk [27]. Likewise, apple pickers spend most of their harvest time in similar high-risk postures [25]. Repetitive and awkward postures thus emerge as primary ergonomic risk factors across diverse fruit-harvesting contexts.
Tool Design and Handling-Related Load
The characteristics of the harvesting equipment played a critical role in shaping biomechanical demands. According to the recorded tool dimensions (Table 5), the baskets were relatively large (≈40 cm depth) and suspended by ropes and hooks, leading to unilateral carriage on one shoulder. This configuration promoted asymmetrical muscle use and potential imbalance. Rotational-handle prototypes have demonstrated reductions in upper-limb muscle activity, including the flexor carpi ulnaris, extensor carpi radialis, middle deltoid, and upper trapezius, along with lower discomfort levels compared to standard designs [28]. In contrast, shoulder-height load carriage alters gait mechanics, reducing single-limb stance time and increasing step-width variability, which can compromise stability and raise fall risk during transit [29]. These effects are compounded by the handling loads observed in this study (Table 6), particularly during loading tasks averaging ~9.5 kg. Although individual tasks such as loading or stabilizing ladders lasted around one minute, their repetition across cycles likely magnifies fatigue and cumulative musculoskeletal strain.
The ladders used were tall (mean height ~5.9 m), with workers typically climbing nearly seven rungs per cycle (Table 5). Climbing under perturbation increases peak harness forces, underscoring the need for secure foot and hand placement to prevent falls [30]. Tasks performed on ladders, such as tool handling and object retrieval, have also been linked to greater discomfort, especially in the knees and neck, due to limited support and constrained postures [31]. Crutches, averaging over 4 m in length, added another stabilizing requirement that may have contributed to physical strain. Well-designed tools and load-handling systems are therefore critical to minimizing biomechanical stress during manual harvesting.
Spatial Layout and Environmental Demands
Workers in this study frequently traversed over 30 meters between the collection point and the trees, with similar return distances, leading to cumulative lower-limb fatigue. These horizontal demands, combined with the vertical reach required to harvest lychees at a mean height of 3.5 meters, imposed substantial physical strain. To access distant or obliquely positioned fruit clusters, workers often elevated their arms and adopted awkward postures, increasing cervical and shoulder loading.
Similar ergonomic constraints have been documented in other fruit-harvesting settings. Orchard configurations such as narrow row spacing and uneven tree alignment limit movement and promote trunk twisting, side-bending, and overhead reaching [25, 27]. These spatial characteristics act as upstream ergonomic risk factors by shaping task execution and body mechanics. Previous reviews further implicate field design variables, including row width, tree placement, and travel distances, as contributors to repetitive movement and postural strain, thereby elevating ergonomic risks in agricultural labor [7, 32].
Task Durations and Repetition of Load
Quantitative video analysis showed that repetitive movements occurred frequently within each harvest cycle, averaging one to three ladder ascents, twelve bunch-breaking actions, and ten dropping actions per cycle (Table 4). With a mean cycle duration of 4.07 ± 1.31 minutes (Table 2), these short yet high-frequency tasks offered limited opportunity for muscular recovery. Comparable short-cycle patterns have been reported in apple harvesting, where picking sequences typically last 3–5 minutes and involve 11–14 repetitions per minute [18, 21, 25, 33]. Similar repetition and sustained exertion have also been documented in oil-palm and blueberry harvesting, where piece-rate systems drive continuous upper-limb activity for 6–8 hours per day with minimal rest [24, 27]. Such frequent repetition and restricted recovery periods contribute to cumulative musculoskeletal loading and elevate MSD risk, even when individual tasks appear brief [21, 24, 25, 27].
This high repetition rate was accompanied by considerable physical load during individual handling tasks. Although individual handling tasks lasted a mean of about one minute, the measured loads were not trivial (Table 6). For example, basket loading required lifting ≈9–10 kg, while ladder stabilization involved supporting ≈5 kg. Even shorter actions, such as unloading baskets or holding crutches, added additional weight-bearing demands. When repeated across multiple harvest cycles, these short-duration but load-intensive tasks resulted in considerable cumulative fatigue. Repetitive handling of tools and produce, often without mechanical aid, increases the risk of overuse injuries. Even short agricultural tasks, when repeated under load, impose substantial biomechanical stress that predisposes workers to musculoskeletal disorders [1]. Such repetition has been linked to disruptions in joint coordination, which may serve as early indicators of fatigue accumulation and elevated MSD risk [34].
Biomechanical modeling using motion capture, surface electromyography (sEMG), and inverse dynamics has demonstrated potential for estimating spinal muscle load and fatigue during repeated lifting, even without directly measuring fatigue progression [35]. Furthermore, task repetition, duration, and posture interact to amplify ergonomic strain in manual agricultural work [36]. Repetition and accumulated task duration should be integral to ergonomic evaluations in physically demanding agricultural work.
Although rest periods were not formally recorded in this study, field observations indicated that most of the workers completed several consecutive harvest cycles before taking short, unscheduled breaks. Evidence from a previous study in the same lychee-harvesting population showed that the total daily rest duration had a mean of approximately 43 minutes (range: 0–120 minutes), with nearly 90% of workers resting less than 20 minutes in total throughout the workday [12]. Such limited overall recovery time, combined with high repetition rates and load-intensive tasks, likely exacerbates cumulative musculoskeletal fatigue and increases the risk of MSDs.
Implications for Ergonomic Interventions
This study highlights the need for targeted ergonomic interventions in lychee harvesting. Priority measures may include redesigning baskets for bilateral carrying, enhancing ladder safety features, introducing mechanical assistive tools, and reorganizing workflows to reduce repetitive strain. Rotational-handle baskets, for example, significantly reduce upper-limb muscle activity and discomfort [28]. Mechanical aids such as the Ergo Bucket Carrier and Easy Lift have demonstrated effectiveness in reducing biomechanical load during manual handling [37]. Mobile platform systems also offer a viable alternative to ladders, reducing overhead reaching and trunk flexion without increasing overall postural strain [38]. However, platform use may raise arm repetition rates, indicating the need to jointly optimize both equipment and task organization [33].
Insights from the Hierarchical Task Analysis (HTA) further support the prioritization of these interventions. The HTA delineated four major task phases, namely harvesting, transporting, sorting, and loading, with the highest physical demands concentrated in the climbing, bunch-breaking, and basket-dropping sub-tasks. These results highlight the need for ergonomic strategies to address both high-risk actions and transitional points within the harvest cycle, such as shifting between ladders and ground or lifting baskets for transport. Beyond equipment, participatory ergonomic approaches involving workers in workflow redesign have been associated with lower RULA scores and improved satisfaction, underscoring the importance of tailoring solutions to worker characteristics such as age, experience, and task familiarity [39]. Systematic reviews further support the effectiveness of integrated ergonomic strategies, including equipment redesign, workflow adjustment, and user training, in reducing injury risks and physical burden in agriculture [40]. Altogether, the evidence supports the development of tailored, multifaceted ergonomic strategies suited to the specific demands of lychee harvesting.
This study has several limitations that warrant consideration. First, although the PATH method enabled systematic coding of postural risks, the analysis was limited to visual observation and did not incorporate direct biomechanical measurement such as electromyography or motion capture. Second, the study focused only on harvesting and transporting tasks, excluding other stages of lychee production (e.g., sorting, packing) that may also pose ergonomic risks. Third, the cross-sectional design restricts the ability to assess long-term musculoskeletal outcomes or causal relationships. Finally, all of the observations were conducted during a single harvest season in Mae Chai District, Phayao Province (Figure 1), which provides essential geographic context but also limits the generalizability of the findings across other regions or crop cycles.
This study applied an integrated HTA and PATH approach to evaluate ergonomic risks in lychee harvesting. The findings highlight the widespread exposure to awkward postures, particularly overhead reaching, trunk flexion, and unilateral load carrying, exacerbated by large baskets, tall ladders, and spatial layouts requiring long travel distances. Although the individual tasks were brief, their high frequency contributed to cumulative biomechanical strain, especially in the neck, shoulders, and lower limbs.
Effective interventions should include redesigning baskets for bilateral use, improving ladder safety, and optimizing orchard layout. The introduction of assistive tools, such as rotational-handle baskets and mobile platforms, combined with participatory workflow redesign, would offer a practical path to reducing MSD risk and improving worker safety in seasonal fruit harvesting.
We are grateful to the lychee orchard owners and local health volunteers in Phayao Province, Thailand, for their valuable cooperation and support throughout the fieldwork. We also sincerely thank all of the participants who voluntarily took part in this study.
None
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Conceptualization and methodology: Prakasit Tonchoy; Investigation: All authors; Formal analysis: Prakasit Tonchoy, Supakan Kantow, Punyisa Pudpong; Writing – original draft: Prakasit Tonchoy, Phakinee Suta; Writing – review & editing: Supakan Kantow, Pannawadee Singkaew; Visualization: Prakasit Tonchoy, Punyisa Pudpong.
