Evaluation of general-purpose lifters for the date harvest industry based on a fuzzy inference system

https://doi.org/10.1016/j.compag.2007.06.005Get rights and content

Abstract

Most dates are still harvested manually around the world because a mechanized method that covers all the needs is not available. General-purpose lifters of various models are used for date harvesting in some orchards in Iran which is the second largest world producer. In this research the Mamdani Fuzzy Inference System (MFIS) was used to evaluate and classify 10 different lifters on the market to find the most suitable one for date plantations in Iran. Five principal lifter features including working height, length, width, payload and price were selected for comparison. To assign membership functions to each lifter feature, the values of tree spacing, tree yields and tree trunk heights were used. The required data were measured in a survey from 9 orchards in Bam and Shahdad cities (two big producers of dates in Iran). Initial analysis of data showed that two features of all studied lifters including width and payload were in required range for working in date groves, therefore, these features were not included in the fuzzy evaluation system. The lifters were graded by both a human date grower expert and the MFIS. Classified results obtained from the MFIS showed 85% general agreement with the results from the date grower expert. Based on the results, a lifter with 12 m of working height, length of 4.1 m and price of 27 million Iranian Rials (€2000) was found as the most suitable harvester for the studied region. The results showed that any new designs of date harvester for studied region must be able to reach to a working height of maximum 13.5 m. Lifters need to have a length of less than 3 m and must be cheaper than 40 million Iranian Rials (€3000). This technique could be used to evaluate and classify any new date harvesting alternative.

Introduction

Dates are one of the most important fruits in more than 30 countries in the desert regions of the world. The total world date production is approximately 6800 thousand tonnes and Iran produces about 880 thousand tonnes (Iranian Ministry of Agricultural Statistics, 2006Anon., 2006). Over the period of 1999–2001, Egypt, Iran, Saudi Arabia and Pakistan have been producing 61% and Iraq, Algeria, United Arab Emirates, Oman and Sudan 29% of the total world date production (Zaid, 2002). There is a major interest in the mechanization of the date harvesting operation, because most of the harvesting is done manually in Middle East, the origin of this fruit. The most popular palm cultural operations carried out in orchards are: pollination, dehorning, pruning, fruit thinning, bunch bending, bunch bagging, pesticide control and harvesting which is the most labor consuming. Brown (1983) showed that among these operations, harvesting, pollination and pruning are the most labor intensive work accounting for more than 80% of the total production costs.

There are two methods of harvesting the date fruits, traditional and mechanical. Using the leaf bases to climb the tree in Iran, Iraq and Libya is a traditional way of harvesting (Nixon, 1969). Some workers use a belt to secure themselves to the tree. In Africa especially in Algeria people dig holes into the tree trunk to climb it easier (Rohani, 1998). They may hammer pegs into the tree trunk, move to next tree on the leaves, or move on a rope. In mechanized harvesting methods, several systems for moving the workers to the fruit have been investigated. Typical mechanized date harvesting systems are vehicles equipped with a number of long arms, at the end of which a worker could stand in a basket for picking the fruits.

Depending on the date variety and climate, the dates are picked by one of the following ways:

  • 1.

    Bunch cutting is used where all fruits on a bunch ripen simultaneously.

  • 2.

    Bunch shaking is used when all fruits on a bunch do not ripen at the same time.

  • 3.

    Selective hand picking is used for varieties with high quality expensive fruits sensitive to vibration and impact.

During the past decades, labor shortage and hygienic issues for manual harvesting have led to development of mechanical date harvesting systems. Perkins and Brown (1964) used a harvesting system in which three men, a truck driver, a boom operator and a bunch cutter or shaker operate together. Saring et al. (1989) in Israel developed an integrated mechanical system that could harvest the fruits by shaking the tree trunk. A roll-out canvas catching frame was commonly used for collecting the fruit. This method is only suitable for those varieties of dates whose fruits on a bunch do not ripen at the same time. Shamsi (1990) designed a sprocket type climbing machine to harvest dates. The operator sits on a frame and holds on to a handle. He then drives the climber by pushing the pedals with his feet. The study calculations showed that the operator could climb with the machine at a speed of 0.2 m/s. Al-Suhaibani et al. (1988) reported on the design of a date service machine for Saudi Arabia which was built at Silsoe College1. A “U” shape platform on the end of a telescopic boom allowed the worker to reach all of the bunches of dates without any additional movement of the platform. Al-Suhaibani et al. (1993) also reported the field test of the machine. It showed that it could harvest a tree in 21 min which was faster, safer and easier than hand harvesting. Shamsi (1998) designed and developed a tree climbing date harvesting test rig in Silsoe College. This machine climbed up the tree trunk to reach the fruit bunches. The prototype could carry a payload of 100 kg of dates and, considering a field efficiency of 75%, could potentially harvest a tree in 22 min which was 18% faster than the manual harvesting system in Iran.

Most of the dates are still harvested manually in most parts of the world including the Middle East, because a mechanized method that covers all needs is not available. General-purpose lifters of various models are used in some orchards in Iran but date growers do not use many of these machines yet. This research was carried out to evaluate and classify available general purpose lifters to find the most suitable one for date harvesting in this region.

There are two kinds of comparisons, absolute and relative. In absolute methods, each alternative concept is absolutely compared with some set of external criteria. In relative methods alternative concepts are compared with each other using measures defined by the criteria. Ullman (1997) describes evaluation techniques as follows: evaluation based on feasibility judgment, evaluation based on technology—readiness assessment, evaluation based on Go/No-Go screening and evaluation based on a decision matrix. These methods are usually based on the designer's experience and will end with a complicated set of linguistic expressions which are difficult to classify and rank.

Fuzzy logic is a self-learning technique which provides a mathematical tool that can convert the complicated set of linguistic evaluation variables into an automatic evaluation strategy. Fuzzy set theory has been applied to a wide range of applications such as control, image processing, filter design, data clustering, pattern recognition and event classification. Chen and Roger (1994) used a fuzzy method to classify plant structures; they found good agreement between the results from fuzzy prediction and human experts. Verma (1995) developed a fuzzy decision support system to aid decisions related to quality sorting of tomatoes. Tomato quality was described as sensory color, fruit shape, size and firmness. The outputs of fuzzy decision support system predicting quality and the day of the highest quality were very accurate when compared with the data provided by an expert. Classifications of uniform plant, soil, and residue color images were conducted with fuzzy inference systems by Meyer (2004). Nejat-Lorestani et al. (2006) used a Mamdani fuzzy inference system (MFIS) as a decision support system to grade a total of 250 Golden Delicious apples based on size and color. Grading results obtained from MFIS showed 90.8% agreement with the results from the human expert.

The main objective of this research was to establish a fuzzy logic system to evaluate and classify lifters to be used in date harvesting industry according to their important features.

Section snippets

Material and methods

Climbing the tree to reach the fruits is the hardest part of date harvesting operation, therefore, attempts for date harvesting mechanization has been focused on methods of lifting the workers to the fruits. To find the most suitable lifter among existing lifters in Iran for date harvesting industry, the important characteristics of these lifters were classified. To do this, the authors contacted three big manufactures of lifters. Important lifter features were found through catalogs and

Results and discussion

Data analysis of Table 2 shows that 90% and 100% of the trees are more than 3 and 2.5 m away from their nearest tree, respectively. The Table 2 shows that 90% and 100% of trees intra row spacing are more than 3.8 and 3.5 m, respectively. It also shows that 100% of trees rows spacing are more than 4.1 m. Table 1 clears that the maximum lifter width is 2.3 m. Therefore, all lifters can pass through 100% of the trees; so, the lifter width was deleted from the lifter evaluation parameters list. Maximum

Conclusions

In this research the MFIS was successfully applied as a decision support system for evaluation and classification of lifters in the date harvesting industry. The main advantage of a MFIS is to integrate and convert the expert date grower's linguistic technical knowledge into a systematic decision making system. Evaluation results obtained from MFIS showed a good general conformity with the results from the human expert.

Evaluation results showed the Balan Sanat “DML 12” and “EHs 1000” models are

Acknowledgements

This paper is supported by organization of management and programming of Kerman province and center of Excellence for fuzzy systems and its applications at Shahid Bahonar university of Kerman, Kerman, Iran.

References (26)

  • J. Jang et al.

    Neuro fuzzy modeling and control

    Proc. IEEE

    (1995)
  • MathWorks

    Fuzzy Logic Toolbox User's Guide, for Use of the Matlab

    (2004)
  • G.E. Meyer

    Digital camera operation and fuzzy logic classification of uniform plant, soil, and residue color images

    Appl. Eng. Agricult.

    (2004)
  • Cited by (31)

    • Multicriteria decision making for the selection of the best renewable energy scenario based on fuzzy inference system

      2020, Renewable-Energy-Driven Future: Technologies, Modelling, Applications, Sustainability and Policies
    • Sustainability Decision Support Framework for the Prioritization of Hydrogen Energy Systems

      2017, Hydrogen Economy: Supply Chain, Life Cycle Analysis and Energy Transition for Sustainability
    • Enhancing China's energy security: Determining influential factors and effective strategic measures

      2014, Energy Conversion and Management
      Citation Excerpt :

      Membership functions are used to calculate a grade of membership that the variable belongs to. Triangular and trapezoidal functions are usually used in fuzzy logic since they are simple and reduce development and execution time [42]. The mathematical definition of triangular fuzzy number and the standard fuzzy set operations [43,44] are presented in Table 1.

    View all citing articles on Scopus
    View full text