Abstract
Introduction
Oranges are the most demanding citrus fruit in the world. As per FAO 2009 estimate, total cultivation area of orange was about 4.2 million hectares which produced 68.5 million tonnes per year [1]. India is producing huge orange and ranked second in area and third in yield in the world. The harvesting of orange requires 35–45% of total production cost [2]. In the view of the numerous advantages of hand harvesting compared to mechanical harvesting of fruit such as low overhead cost, versatility and adaptability of humans, and visual image processing ability, rapid detection of suitable fruit for harvest, etc. [3], plucking of orange in north-east states of India is being done by climbing on the tree along with a basket. This method of orange harvesting is very risky and hints for sever accidents. Further, this method also leads to awkward postures which increase muscle and joint strain injury [4–6] especially when the orange basket is full or, partially full. Therefore, a high output and comfortable manual hand harvester was essential for the farmers as mechanical harvesters and other mechanised hand harvesting mechanism is not feasible in this region due to hilly terrain and uneven topography.
The limitations of manual harvests are mainly low plucking rate and heavy discomfort in operation, which depends on various factors such as harvester’s design, work environment, physical strength, fruit accessibility, human endurance, etc. The most commonly used parameters for assessment of the physical workload are oxygen consumption and heart rate. Pheasant [7] concluded that the heart rate is most frequently assessed physiological variable in field work than oxygen consumption. Recording of heart-rate responses was made relatively easier with the introduction of heart-rate monitors. Several subjective (psychophysical) measures have also been proposed for quantifying the physical workload experienced by workers while performing various types of activities. Borg’s rating of perceived exertion (RPE) [8] and body part discomfort score [9] are most commonly used psychophysical methods in ergonomic evaluation for subjective assessment of discomforts. Borg scale or RPE is not a measure of responses for individual body segments, but rather a subjective judgment about the task and its effects on the body as a whole [10, 11]. Helm and Heimstra [12] found a high correlation between perceived workload and task difficulty. Awkward working posture is one of the most responsible factors for prevalence of work related musculoskeletal disorder. Intensive research around the world has focused on improving the workers’ performance in various fields and bringing down their cost. Since manual fruit harvesting is highly physical demanding task and it requires considerable amount of energy. Therefore, it is essential to evaluate the physical workload involved in manual orange harvesting.
Keeping the above facts in mind, present study was formulated to develop manual orange harvester considering the hitches of existing harvesters, and compare the performance of the developed harvester with the existing harvesters which was being used for orange harvesting.
Methodology
The methodology of this study included the preliminary study of two existing hand operated manual orange harvester and development of a new model by eliminating the problems associated with the existing harvesters. The developed model along with existing models was evaluated extensively. The evaluation was in two stages: Laboratory and field experiment. In the laboratory a relationship between heart rate (HR) and oxygen consumption was established to estimate energy expenditure rate (EER), because measurement of oxygen consumption in the field condition is tedious and time consuming as stated above. In the field, heart rate of the subjects was measured and oxygen consumption was predicted from the established relationship. Further, in the field experiment, performance parameters of orange harvester i.e. plucking rate (PR), damaged quantity (DQ), plucking energy requirement (PER) and discomfort rating were also observed.
Existing harvesters
The two most commonly used orange harvesters were purchased named as existing model 1 (EM1) and existing model 2 (EM2) which are shown in Fig. 1.
The EM1 consists of a cutter to cut the orange and a basket on top of the harvester to collect the harvested oranges. This model was relying on rope pulling mechanism to actuate the cutter. The harvester has to be brought down after plucking every 4-5 fruits for unloading it. The following four major problems were perceived during the operation, resulting in lower efficiency and acceptance of the harvester among the farmers. The insertion of cutter in fruits’ stem was difficult and time consuming due to the smaller size of cutter’s nose and improper design of the cutter angle. During rope pulling, the harvester has to be hold by one hand only as other hand was required for rope pulling which was resulting in disturbing the selected position of the cutter. This process was also found to be causing in wrist pain of the user as s/he had to hold the same with one hand during the process. The fruits collecting baskets create heavy moments with the accumulation of fruits in it, thereby resulting in added pain in the wrist, arm and shoulder of the worker. Further, quite frequently fruits fall outside the basket and results in fruit damage.
The EM2 has cut and hold type mechanism. It consists of clutch for actuating the cutter and telescopic handle for varying length of handle. On the operation of the clutch after proper insertion into the stem, fruit hold in the cutter and harvester has to be brought down to release after each and every fruit. Therefore, this model did not require basket to collect the harvested fruit. This was light in weight and hence comfortable, but time consuming.
Development of harvester
Taking due considerations of the above problems in mind, a need based modified orange harvester was developed (Fig. 2). The specification of the developed model (DM) is tabulated in Table 1 along with EM1, and EM2 for comparisons. The major highlights of the developed model are as follows: A long-nose cutter at an angle of inclination 280 with the line of harvester’s handle (Fig. 3a) was provided to solve the problem of insertion of clutter into fruits’ stem as observed in EM1 (Fig. 3b). A clutch system was incorporated to actuate the cutter which did not require to leave the harvester to operate the cutter. A continuous net from basket to the ground was provided for instant unloading of harvested fruits slowly without consuming extra time. This also eliminated the additional moment on the wrist due to accumulated harvested fruits at the top. The basket at the top adjusted in such a way to ensure that all the harvested fruits should fall into the basket.
Selection of subjects
Twenty healthy male and female experienced subjects were selected for this study. All subjects were right-handed. Initially, subjects were screened and excluded from the study if they had reported any previous or current musculoskeletal disorder or low back injury at the time of experiment. The subjects were informed about this study and taken their consents of participation before the experiments. They were also asked to refrain from heavy exercise for three days prior to the data collection process. The body surface area (BSA) in square metres was calculated from the subject’s height and weight by the formula of DuBois and DuBois [13] using Eq. 1. Further, body mass index (BMI) was computed using weight and height parameters (BMI = weight (kg)/Height (m2). The BMI was classified as underweight (BMI <18.5), normal (BMI = 18.5 to 24.9), overweight (BMI = 25 to 29.9), and obese (BMI >30) according to the WHO [14]. The mean BSA and BMI were derived from individual BSA and BMI data.
The cardiorespiratory response of all the selected subjects was observed in the laboratory.
The relation between the oxygen consumption rate and HR was established for each subject while working on a treadmill (Trackmaster, TMX425) using an ambulatory instrument (Cortex, MetaMax® 3X) before conducting the field experiment (Fig. 4).
This relationship is relatively easy and cost-effective thereafter to quantify HR in the field and predict VO2 by interpolation [15]. Prior to the session, MetaMax 3X equipment was calibrated for volumetric calibration. The gas analyzers were calibrated initially against ambient air and then 16.10% O2, 4.90% CO2 and 79% N2 mixture. A typical relationship between HR and oxygen uptake of a subject is shown in Fig. 5. The HR measured during orange harvesting was used to calculate VO2 by the calibration curve of the respective subjects. The energy expenditure rate was determined by an established relationship between oxygen consumption and energy expenditure [16] as given below(Eq. 2).
where, EER = Energy expenditure rate, kJ/min and
OCR = Oxygen consumption rate, l/min.
The field experiment took place in Arunachal Pradesh, India. The experiment was designed to collect the performance parameters of manual orange harvesters. All related operations were executed in 7 different orchards. The experimental area was mostly sloppy ranging from 5–25%, hence use for machinery is not feasible and suitable.
Heart rate measurement
Heart rate (HR) increases as physical workload and energy demand increases. During moderate work intensities, HR is reasonably linearly related to oxygen consumption [17, 18]. The HR was measured using polar heart rate monitor (Polar, RS400). Before fixing the transmitter it was wiped-off with the spirit using cotton to remove dust and sweat of the previous subject. After that, transmitter of the heart rate monitor was fixed on the chest and receiver (wrist watch) was tied on subject’s preferable hand. The HR recording interval was chosen for 60 s to avoid numerous reading [19]. Before recording the HR, the subject was allowed to relax for 5 minutes silently. The time of start of the test was recorded on the data sheet. Mean values of physiological responses during last 5 minutes were taken as the resting values for each participating subjects. The subject was asked to pluck oranges after rest for the duration of 60 minutes. At the end of the experiment, the subjects were allowed to sit in relax position until the recovery HR equalled the resting HR. The recorded data were transferred to a computer through S-series infrared interface and analyzed resting, working and recovery heart rate.
Performance parameter calculation
The performance of orange harvesters were evaluated by plucking rate (PR), damage quantity (DQ), energy expenditure rate (EER), plucking energy requirement (PER) and discomfort rating (DR). The PR was calculated by counting the total numbers of oranges plucked per hour. Similarly, DQ was calculated by counting total number of damaged fruits in 100 pieces of harvested orange. The EER was calculated by using Eq. 2 for subsequent prediction of VO2 from the measured HR. The EER was divided by the plucking rate with consideration of unit factor to find out the PER.
Discomfort level scale
Comfort and discomfort have been considered as two discrete states, opposites on a continuous scale ranging from extreme comfort through a neutral state to extreme discomfort [20]. The subjective measurements are one of the important methods in measuring the discomfort or comfort level in the respective user [21]. The perceived regional discomfort for each subject was assessed after 60 minutes for each of 8 body regions (Fig. 6) i.e. neck, shoulders, upper arms, lower arms, wrists, thighs, knees and lower legs. Each body region was numbered differently to avoid confusion. The discomfort score was given by the Borg’s CR-10 scale. The scale anchors were as follows: (nothing at all, 0; very slight, 1; slight (light), 2; moderate, 3; somewhat severe, 4; severe (heavy), 5; very severe, 7; and very, very severe, 10). Subjects select a number from 0 to 10 based on their perception of discomfort. The subjects were thoroughly familiarized with the assessment techniques prior to participation in the study.
Results and discussion
Physical characteristics
The body weight and height were measured to the nearest of 0.1 kg and 0.1 cm using a standard balance and integrated composite anthropometer (IIT Kharagpur, India), respectively, with subjects wearing light clothing and without shoes. The subjects were ranged in age (year), stature (cm), weight (kg) and work experience (year) from 22–46, 151–167, 40–64 and 5–12 for male and 21–38, 147–159, 42–58 and 5–10 for female, respectively. A summary of the subjects’ physical characteristics is presented in Table 2. Two subjects (female) were underweight, one subject (male) was overweight and rest seventeen subjects were in the normal BMI range (18.5–24.99). Average BMI (±SD) was found to be 23.08 (±2.81) for male and 19.04 (±1.29) for female.
Performance evaluation of orange harvesters
Performance of EM1, EM2 and DM orange harvesters were evaluated in terms of plucking rate (PR), damaged quantity (DQ), energy expenditure rate (EER) and plucking energy requirement (PER) which are represented graphically as shown in Fig. 7. In functional aspects, the overall performance of the DM was found to be better than the two existing models (EM1 and EM2) because of all the major issues of existing models were addressed in the DM. The plucking rate of the DM was 425 pieces per hour compared to 300 and 287 for EM1 and EM2, respectively. It was evident that, the DM did not require additional effort and time for unloading the plucked fruit which was proven reason of higher plucking rate.
Unloading of fruits in EM1 was required after every 4–5 pieces of harvested fruit, however, in EM2 it was required after every piece of harvested fruit which was consuming more time and subsequently more discomfort and fatigue. Further, ease of picking the cutting position by providing suitable changes in cutter enhanced the work output.
The damage of fruits was also one of the major concerned while evaluating the performance of harvester [22]. The damage of fruits was mainly caused by outside fall of fruits from the collecting basket just after the cut. Further, if the fruit did not cut along with a small stem that was also considered as damage because the self-life of such orange reduces significantly. The minimum damage was observed in the DM compared to other models due to clear visibility of fruits’ stem, provision of clutch, inclusion of the long nose cutter at the appropriate angle and proper alignment of the basket with cutter.
Further, overall performance of all models was compared by plucking energy requirement (PER) per unit fruit. It was observed that energy expenditure rate (EER) in EM2 was found minimum (14.1 kJ/min) because of light and handy harvester followed by DM (15.2 kJ/min) and EM1 (20.1 kJ/min) but plucking rate was different which resulted in different PER. The PER was observed minimum in case of the DM (2.14 kJ/piece) followed by EM2 (2.95 kJ/piece) and EM1 (4.02 kJ/piece). Hence, overall performance of the developed model in terms of PER, DQ and PR was found better than the other two existing models.
Musculoskeletal discomfort
The whole body part discomfort and local body part discomfort experienced by each subject were analyzed. The developed model was scored lower discomfort (score:<3.5 out of 10) followed by EM2 (score:<4.1 out of 10) and EM1 (score:<4.6 out of 10). Further, discomfort reported for neck, shoulders, upper arms, lower arms, wrists, thighs, knees and lower legs for each model were analyzed separately. Rating of body parts discomfort as reported by the subjects indicated that neck, shoulders, upper arms and wrists had higher discomfort compared to other body parts (Fig. 8).
Kee [23] also reported an increase in discomfort rating caused by increased upper arm elevation. The quantification of postural movements undergoes the influence of several factors, such as the height of trees, and the use of harvesting equipment and tools [24]. The orange harvesting process involved hold, cut and pushing actions with the help of both the hands in standing posture. Hands, shoulders, upper arms, lower arms, palms, and wrist were involved in this action. Since the subjects need to move frequently from one place to another place, therefore less discomfort reported in knees and lower legs. Further, pains are reported in shoulders, upper arms and lower arms because about 40 to 50% of total time spent by workers keeps their hands over the shoulder [5, 6].
Conclusions
In this study, a manual orange harvester (DM) was developed by eradicating various functionality problems encountered in two commercially existing harvesters (EM1 and EM2). The findings of the results showed substantial decrease in physiological cost and damage quantity with increased work output in the developed harvester compared to existing harvesters. The energy requirement per unit plucking was found 47 and 27 percent less in the DM compared to EM1 and EM2, respectively. Similarly, the plucking rate in the DM was found 42 and 48 percent higher than the EM1 and EM2, respectively. Therefore, the developed model was found much better than the existing harvesters and well appreciated and accepted by the farmers. Further evaluations are necessary in various hilly regions of country with larger sample size to conclusively establish the benefits of long-nose cutter, clutch system and instant unloading mechanism of the developed harvester.
Limitations of the study
The developed model is suitable for hilly region where hand plucking is difficult. Developed harvester is suitable for fruit harvesting of orange trees up to 3 m height. Harvester’s handle is fixed in length. A telescopic handle may give better performance. Evaluation was carried out in only one selected region of northeast India i.e. Arunachal Pradesh with limited number of subjects/participants.
Conflict of interest
None to report.
Footnotes
Acknowledgments
The authors are grateful to the Indian Council of Agricultural Research, New Delhi, India for financial support for conducting this study. Authors would also like to express their gratitude towards Dr. L. P. Gite for his valuable guidance throughout the study.