Next Article in Journal
Effect of Nutrient Management on Soil Carbon Quantities, Qualities, and Stock under Rice-Wheat Production System
Previous Article in Journal
Effects of Extraction Solvents on the Total Phenolic Content, Total Flavonoid Content, and Antioxidant Activity in the Aerial Part of Root Vegetables
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Agriculture
Volume 12
Issue 11
10.3390/agriculture12111821
agriculture-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

De-Handing Technologies for Banana Postharvest Operations—Updates and Challenges

by
Jie Guo
1,2,3,†,
Jieli Duan
1,2,*,†,
Zhou Yang
1,2,4 and
Manoj Karkee
3
1
College of Engineering, South China Agricultural University, Guangzhou 510642, China
2
Guangdong Laboratory for Lingnan Modern Agriculture, Guangzhou 510642, China
3
The Center for Precision and Automated Agricultural Systems, Washington State University, Prosser, WA 99350, USA
4
School of Mechanical Engineering, Guangdong Ocean University, Zhanjiang 524088, China
*
Author to whom correspondence should be addressed.
†
These authors contributed equally to this work and share first authorship.
Agriculture 2022, 12(11), 1821; https://doi.org/10.3390/agriculture12111821
Submission received: 4 October 2022 / Revised: 23 October 2022 / Accepted: 27 October 2022 / Published: 1 November 2022
(This article belongs to the Section Agricultural Technology)
Browse Figures
Versions Notes

Abstract

:
Many aspects of the agricultural industry such a field crop planting and harvesting and chemical application in fruit crops have been employing mechanization and automation solutions for decades. However, the de-handing operation in banana postharvest operations is usually performed manually. Mechanical or automated de-handing is a potential long-term solution to address labor shortages and the associated high costs. Bananas are mainly grown in developing countries located in tropical and subtropical regions, where the development of agricultural mechanization and automation solutions started only recently and is progressing relatively slowly. In addition, large-scale banana orchards are mainly distributed in hilly and mountainous areas, though there are also some small-scale banana plantations in plain areas. The complex environment of banana orchards and the aging farming population are other important factors that make it difficult to realize mechanized operation of banana de-handing. In recent years, researchers have proposed advanced techniques that may facilitate the development of mechanical de-handing systems. However, the successful adoption of mechanical de-handing technology still faces many challenges. This paper systematically reviews the existing research on de-handing technologies and component mechanisms. A comprehensive evaluation is carried out from the perspectives of feasibility of the mechanism design, stability of the model simulation and reliability of the prototype systems developed. The future challenges and opportunities for designing and practically adopting mechanical de-handing equipment are also summarized and discussed.

Graphical Abstract

1. Introduction

Bananas are the largest fruit crop in terms of planting area and trade volume in the world. Banana planting is mainly concentrated in developing countries in tropical and subtropical regions such as China, Brazil, India, Ecuador and the Philippines [1,2] where it provides one of the main sources of income for farmers [3]. The harvest and postharvest operations in bananas primarily include field picking, ropeway transportation, de-handing, cleaning and disinfecting, and packaging and distributing to grocery outlets (Figure 1). With the development and advancement of mechanization, automation and intelligent technologies, the production efficiency of banana farmers has been greatly improved. Currently, various operations in banana postharvest operations have been mechanized to varying degrees; however, the de-handing operation still relies on manual labor [4].
With the recent societal and economic changes including urbanization and wider commercial activities in urban areas, young and middle-aged laborers are rapidly leaving the farming communities in the countryside and entering larger cities, which leads to aging of the rural population thus creating increasingly serious challenges for farming. Therefore, availability and cost of manual labor has become the most critical issue for banana farmers, especially for the time-sensitive and labor-intensive operations. There are also multiple challenges in retraining workers with operational skills and de-handing experience. To address these challenges and improve long-term sustainability of the banana industry, mechanical de-handing offers a potential alternative to manual de-handing [5]. Banana planting is mainly concentrated in hilly and mountainous areas with uneven terrain and steep slopes, which are not conducive to pedestrian activities. In addition, in order to improve returns to farmers, banana farmers have adopted different planting patterns in different regions, as shown in Figure 2. In the context of the integration of agricultural machinery and agronomy/crop architecture, the complex geographical environment and diverse planting patterns in banana orchards have posed additional challenges for the design and application of de-handing machines.
The purpose of this paper is to present the latest literature on banana de-handing machinery in order to assess the current situation and to find potential solutions to this critical technical and scientific challenge. The organization of this paper is as follows: in Section 2, an overview of de-handing methods including traditional manual work patterns is provided. In addition, existing key mechanical de-handing techniques have been reviewed. In Section 3, potential solutions to existing challenges are proposed in four different aspects: the development of a visual perception system, the design of a cutting tool with profiling function, the exploration of de-handing mechanisms and eye-hand synergy and the evaluation of mechanical damage of banana hands after de-handing. Finally, in Section 4, general conclusions of the review are summarized.

2. Overview of De-Handing Methods

2.1. Traditional De-Handing Methods and Challenges

Banana de-handing is the action of separating the banana hand (fruit) from the bunch stalk (stalk). Traditionally, workers use a simple cutting tool to cut the node one by one along the central axis of bunch stalk from the first hand to the last/end hand to complete the task of de-handing whole banana bunches, as shown in Figure 3. A banana bunch has complex morphological features. The banana hands grow and distribute nonlinearly in both the axial and circumferential directions of the bunch stalk [6]. The overall quality of the bunch is varied and the fruit is easily mechanically damaged.
Manual de-handing is labor-intensive and physically challenging. Not only does the de-handing quality vary due to workers’ skills and experience, but repetitive motion in a small operating space will cause potential hazards to workers’ health that cannot be ignored [7]. In addition, some de-handed banana hands do not meet the quality requirements for fresh market consumption, and therefore they are usually subjected to secondary cutting, which inevitably increases the labor use and production cost. The current manual de-handing mode, therefore, hinders the sustainability of banana postharvest operation and the entire industry. The banana industry urgently needs to transform from the traditional de-handing mode/technique to a modern mechanical de-handing operation.

2.2. Mechanical De-Handing

There have been some efforts in developing mechanical de-handing techniques to replace manual work. In order to effectively connect the cableway transportation and the cleaning of the banana hands, the mechanical de-handing equipment usually includes the clamping mechanism of bunch stalk, the lifting mechanism for the bunch, the de-handing mechanism, the collection mechanism for banana hands and the crushing mechanism for bunch stalk, as shown in Figure 4. Before working, the banana farmer places the de-handing machine at the bottom of the ropeway, hangs the banana bunches picked from orchards on the transportation ropeway and gradually approaches the de-handing machine. The position of the bunch closest to the de-handing machine is slightly adjusted, so that its central axis basically coincides with the center of the machine. Then the clamping mechanism is adjusted to the top of bunch stalk, and the de-handing mechanism to the first banana hand at the bottom of bunch stalk. Finally, the liftable legs of the de-handing machine are adjusted to locate the flexible conveyor belt below the de-handing mechanism, so as to complete the preparatory work for the de-handing operation.
When the banana bunches hanging on the transportation ropeway are transported from left to right, the clamping mechanism clamps the bunch stalk, and then the de-handing mechanism completes the cutting operation of the first banana hand. In the process of de-handing from bottom of the bunch to top, the diameter of the bunch stalk gradually increases, the profiling cutters of the de-handing mechanism move radially to adapt to the changing diameter of the stalk, and then the de-handing is complet4d hand by hand in the bunch. The remaining bunch stalks fall into the cutting and pulverizing mechanism at the bottom of the de-handing machine and are cut and pulverized on the spot, which is convenient for the subsequent unified operation of returning the stalks to the field. After the de-handing work is over, the flexible conveyor belt and the collection mechanism of banana hands are put away, and the de-handing machine is moved to the workshop to prepare for the next job.
The de-handing mechanism is an important part of the de-handing machine and it is the core of realizing the banana de-handing operation. Since the diameters and curvatures of the same bunch stalk where banana hands grow are different, the above parameters are also different between different bunch stalks. Therefore, self-adaptive profiling performance of the cutting tool in the de-handing mechanism, with regards to a bunch stalk with irregular geometric shapes, directly determines the success rate of de-handing and the incision quality of the banana hand, and also indirectly affects the smoothness of the de-handing operation and the fruit quality of the banana hand. The chopped banana hands and the fruit with mechanical damage cannot be stored and sold for fresh market consumption. Therefore, this paper provides a detailed analysis and evaluation of the mechanical de-handing cutters and equipment that have been reported so far, and recommendations for future research will be made based on the findings of these studies.

2.2.1. Circumferentially Rotating Mechanical De-Handing

The banana hands are distributed in the circumferential and axial directions on the bunch stalk; the circumferential angles are different for different bunches and among different banana hands on the same bunch [6]. Based on this hand configuration, mechanical de-handing methods have been investigated based on circumferential rotation along the bunch stalk. It is noted that the performance of these de-handing mechanisms is assessed based on how well the mechanism avoids cutting into the bunch stalks and banana hands, and avoids damaging the banana fingers.
Referring to the working principle of aircraft engines in the aerospace field, Yang et al. designed a banana de-handing cutter with an annular shape and a circle with variable enveloping diameter, as shown in Figure 5 [9]. This de-handing cutter included double-edged cutting blade, annular knife holder, connector and an A-shaped tie rod mechanism. The cutting blades are installed on the guide groove of the annular knife holder to form an annular enveloping structure capable of wrapping the bunch stalk. The cutting blades were staggered and each blade was divided into upper and lower parts, the upper part a vertical blade body with double edges on the top, and the lower part an outwardly inclined adjustment. Through the adjusting part, the annular enveloping structure can be contracted inward or expanded outward, so as to realize the function of adaptively varying the diameter. In addition, the distance between the annular knife holder and the central axis can be changed by the swinging of the A-shaped tie rod mechanism, thereby ensuring that the shape of the annular enveloping structure is circular or approximately circular. The de-handing cutter can automatically adjust the diameter of the annular enveloping structure according to the change of the diameter of bunch stalk, and can avoid missing cutting nodes.
Xu et al. designed a circular arc profiling mechanism based on the working principle of the expandable mechanism, and manufactured a circumferentially expandable banana de-handing platform on this basis [10]. The circular arc profiling mechanism they proposed was composed of upper, middle and lower layers of plane slides. The upper layer of the mechanism was composed of an expandable arc, arc support rods and driving slides; the middle layer of the mechanism was composed of fixed slides, linear guides, push rods and transmission rods; the lower layer was composed of adjusting slides and linear guides. Figure 6 shows the schematic diagram of the arc profiling mechanism and the de-handing platform. This study showed that the de-handing platform has good profiling accuracy (>96.5% within the profiling range of the bunch stalk). In addition, the incision quality of the banana hand after de-handing is greatly improved compared with that of the manual operations. This research is significant in guiding the de-handing operation of a single banana hand. However, in commercial banana orchards, banana bunches are usually required to be de-handed continuously throughout the whole bunch to achieve the desired throughput for practical adoption, which could be a direction for future advancement of this technique.
Duan et al. also designed a de-handing platform with self-adaptive function for the banana bunch stalk, which primarily consists of a cutter plate, slider, V-shaped connecting rod, cutting blade, Hook-like hinge, motor and frame, as shown in Figure 7 [11]. Through the rotation of the Hooke-like hinge, the cutter plate and the cutting blade of the de-handing platform can rotate by a corresponding angle in space, so as to adapt to the change in the curvature of the bunch stalk. Since the rotation angle is controlled by motors, the profiling effect is more precise. In addition, the cutting blade can adjust its expansion area according to the circumferential angle of the growth of the banana hand, so as to complete the cutting operation of banana hands with different circumferential angles. This feature of the de-handing mechanism substantially improved the performance, which was valuable in addressing: (i) the poor profiling effect of existing de-handing mechanisms on the curvature of bunch stalk; (ii) low cutting accuracy of the cutting blade in the circumferential direction.

2.2.2. Mechanical De-Handing with Axial Plunge-Cutting

In addition to the circumferentially rotating mechanisms discussed above (Section 2.2.1), a number of studies have reported on the axial plunge-cutting mechanism for mechanical de-handing along the bunch stalk. There is a big difference between the two mechanical de-handing methods. The former realizes the de-handing operation of a single banana hand according to the different circumferential angles of banana hands growing on the bunch stalk. The research focuses on the circumferential profiling design of the cutting tool. The latter is based on the characteristic that the banana hands are staggered along the axial direction on the bunch stalk, and realizes the de-handing operation of all banana hands one by one for the whole bunch. The research focuses on the self-adaptive performance of de-handing equipment to changes in the curvature and diameter of the bunch stalk. Considering the various requirements of the actual de-handing operation in banana orchards, we found that the axial plunge-cutting mechanical de-handing method along the bunch stalk has more potential for long-term development.
Yang et al. designed a stalk diameter-adaptive de-handing mechanism consisting of frames, motor, cutter plate, gear set, connecting rods, linear guides, L-shaped slider, springs and cutting blades, as shown in Figure 8 [12]. The gear set is installed under the frame and consists of a driving gear and a corresponding mesh. The linear guides are arranged on the cutter plate in a circular array with the center of the cutter plate as the center of the circle. Each blade holder can move in the radial direction with the varying diameter of the bunch stalk, so as to adjust the expansion area of the annular envelope and profile the bunch stalk. By introducing the crank-slider mechanism, the six cutting blades evenly distributed around the circumference can move synchronously in the radial direction using the gear drive, which provides a reference for the design of the cutter, self-adapting to the varying diameters of the bunch stalk.
Yang et al. designed another plunge-cutting de-handing mechanism with variable diameter, as shown in Figure 9 [13]. This mechanism primarily included cutting blades, cutter plate, synchronous control mechanism for the multiple cutters, air cylinder and twistable annular hoop. The cutter group synchronization control mechanism is arrayed on the cutter plate with the center of the cutter plate as the center, and is mainly composed of L-shaped supports, U-shaped blocks, connecting rods, incomplete gears, sliders, vertical guides and springs. During the de-handing process, the annular blade that wraps the bunch stalk moves radially under the drive of the synchronous control mechanism of the cutter group. Once a hand is de-handed, the bunch stalk at the center of the mechanism is moved down manually or mechanically, and the expansion area of the annular envelope is adjusted by the joint action of the torsional deformation of the twistable annular hoop and the deformation mechanism, so as to achieve the purpose of self-adapting to the varying diameters of bunch stalks.
Yang et al. proposed a wire-cutting mechanism for de-handing that consisted of, primarily, a cutter plate, linear guide, L-shaped slider, constant force coil spring, cutting blade and retractor with metal wire (Figure 10) [14]. The retractor can be installed between any two linear guides and fixed on the cutter plate, and it can quickly and automatically retract or extend the wire. The cutting blades are connected in series by a metal wire. The cutting blades and the metal wire form an annular envelope, and the metal wire between adjacent blades is used for cutting. For a gradually increasing stalk diameter, the L-shaped slider moves radially, and the length of the metal wire between the cutting blades is elongated to adapt to the change of the diameter of bunch stalk. The cylinders installed under the cutter plate provide power to move the cutter plate upward as a whole. The metal wire in the retractor is tensioned at the moment when the blades cut the node, and finally the blade and the metal wire complete the de-handing operation together. When the whole bunch is completely de-handed, the bunch stalk is removed, the L-shaped slider is reset under the action of the coil spring and the excess metal wire is automatically recovered into the retractor. Using the method of combining the cutting blade and the metal wire, the gap formed by the blade during the radial movement can be made up to ensure that there is always a metal wire between the adjacent blades, so that the banana node will not be missing. Retractors with balls and coil spring allow for rapid wire tensioning and retraction, reducing blade recovery time and improving work efficiency. Therefore, this mechanism minimizes or avoids issues such as short radial expansion stroke of the whole blade, large cutting gap between adjacent blades and long recovery time of blades.
In order to better connect the de-handing operation with the cableway transportation operation and the cleaning operation of banana hands, Yang et al. designed an iris-gear-like mechanism for de-handing based on the working principle of the iris mechanism [15]. This de-handing machine was primarily composed of a clamping mechanism for the bunch stalk, a lifting mechanism, a de-handing mechanism, a flexible conveyor belt with liftable brackets and a crushing mechanism, as shown in Figure 11. This mechanism can realize the radial opening and closing of the arc blades, thereby improving the self-adaptability to the varying diameters of the bunch stalks. In addition, the de-handing machine also has the functions of collecting banana hands and crushing bunch stalk. The de-handed banana hands can be safely transported away by the flexible conveyor belt, which reduces the mechanical damage to the fruit. The remaining bunch stalk is crushed in-situ, which is convenient for the subsequent stalk returning operation. This design can effectively integrate the overall operations in a banana postharvest environment.
Yang et al. also designed a de-handing mechanism in which the cutting blades can move synchronously in the radial direction and vertically in the axial direction [16]. Its core components included lifting mechanism, fixed center and variable diameter mechanism and cutting blade group. Before starting the de-handing operation, the banana farmer fixes the end of the bunch vertically in the working space of the cutting blade group, and ensures that the center of the bunch stalk and the center line of the cutter plate coincide. The motor drives the entire de-handing plate to raise it slowly and then the node is severed by the cutting blade group. The schematic diagram of this de-handing mechanism is shown in Figure 12. This mechanism was able to de-hand one hand at a time or de-hand the whole bunch continuously without external labor or devices to lift the bunch. In addition, the self-centering and adapting mechanism for variable-diameter stalks can realize the synchronous radial expansion or contraction of the cutting blade group. The staggered connection form is adopted between adjacent cutting blades, so that the blades can always be effectively connected during the radial movement. This design ensures that there is no missed cutting phenomenon in the de-handing process in the circumferential direction, thus improving the de-handing efficiency and cutting quality.
Guo et al. designed a de-handing mechanism based on the working principle of the expandable mechanism widely used in the aerospace industry, which is composed of a cutter plate, linear guides, sliders, cutting blades, connecting rods and constant force components [17]. The schematic diagram of this mechanism is shown in Figure 13. Two cutting blades and two connecting rods form a parallel four-bar linkage mechanism, a plurality of parallel four-bar linkage mechanisms form a symmetrically shaped expandable mechanism and the radial extension or retraction of the expandable mechanism is achieved by sliders. When the diameter of the bunch stalk varies, the slider moves radially on the linear guide, and the diameter of the envelope circle formed by the annular cutting blade group changes synchronously with the diameter. Each slider corresponds to a constant force component, which is used to ensure that the clamping force of the annular cutting blade group on the bunch stalk remains constant during the radial movement. The automatic retraction function of the constant force coil spring can also help the cutting blade group to retract and reset after completing de-handing. This de-handing mechanism accommodated a stalk diameter range of 40 to 84 mm and was suitable for continuously de-handing entire bunches.
According to the working principle of the three-spring parallel mechanism, Yang et al. and Guo et al. designed a de-handing mechanism, which consisted of cutter plate, torsion springs, horizontal compression springs, uprights, inclined compression springs and cutting blades, as shown in Figure 14 [8,18]. The cutter plate, cutting blade and frame formed a two-degree-of-freedom series compliance mechanism. The two cutter plates that were rotated around their fixed axes released the rotational freedom in two directions, which ensured self-adapting to the varying curvature of bunch stalk. The torsion spring could be adjusted to reduce the impact of the cutter plate and make the rotation of the cutter plate more stable. Under the action of the constant force mechanism, the ring-shaped cutting blade group can achieve a constant clamping force on bunch stalks, thereby making the continuous de-handing process smoother. The experimental results showed that the constant force mechanism and the rotating cutter plates with Hooke hinge in the self-adaptive system are modeled correctly and the mechanism can ensure de-handing quality while self-adapting to variations in bunch stalk. The success rate of de-handing is 100.0% and the pass rate of node incision is 81.3%.

2.2.3. Evaluation of Mechanical De-Handing Technologies

Since there’s no common evaluation system/criteria available currently for the banana de-handing operation, after reviewing the ten banana de-handing technologies and equipment, we evaluated each system/mechanism using three performance measures [8]: success rate of de-handing, incision quality and self-adaptive profiling effect. The design of the circumferentially rotating de-handing cutter shown in Figure 5 draws on the working principle of the axisymmetric vector nozzle, and the ring-shaped cutting blades can be radially expanded or unfolded synchronously to cope with changes in the diameter of the bunch stalk, which is expected to be applied to the continuous de-handing of whole bunches. At present, this de-handing cutter has been authorized by the China National Intellectual Property Administration; however, there is no report on its processing, manufacturing and actual operation. The circumferentially expandable de-handing platform shown in Figure 6 uses a thin blade made by stainless steel as a cutting component, which can be opened or closed in the circumferential direction under the drive of the expandable mechanism. This de-handing platform has better profiling and cutting effects for individual banana hands, and the success rate of de-handing is 100%. However, no studies have been reported with this technique on the continuous de-handing of all banana hands in the whole bunch. The de-handing platform shown in Figure 7 includes an expandable cutting blade to deal with the circumferential growth characteristics of the banana hand and a support base to accommodate the change in the curvature of the bunch stalk. The design of the cutting blade draws on the working principle of the expandable mechanism, and three motors arranged at 90° in space are introduced into the support base to drive the Hook-like hinge. The de-handing platform based on this design may be used as an end-effector of a robot manipulator to complete the de-handing task.
Compared to the circumferential rotating de-handing techniques, axial plunge-cutting-based de-handing techniques have been more widely investigated in recent years, and the related mechanical equipment has been authorized by the China National Intellectual Property Administration. In the de-handing mechanism shown in Figure 8, the introduction of the crank-slider mechanism enables the annular cutting blades to move synchronously in the radial direction, so as to self-adapt to the change in the diameter of the bunch stalk during the de-handing process. This study provided a foundation for further investigating the axial plunge-cutting technique for de-handing. In order to obtain a larger radial motion stroke to improve the self-adaptive performance of the diameter change of bunch stalk, the radial self-adaptive mechanism is optimized in Figure 9. The radial motion of the cutter is converted into the linear motion of the combination of rack-incomplete gear in the vertical direction, which improves the profiling performance of the cutting blades to the bunch stalk. The mechanism shown in Figure 10 introduced a wire-cut de-handing method for the first time, and the balls of different masses have different inertias when moving. Using this principle, and with the help of a coil spring, the tension and relaxation of the metal wire between the cutting blades during the de-handing process are realized. In order to better connect the postharvest operations of bananas, the de-handing machine shown in Figure 11 included the clamping mechanism for the bunch, the de-handing mechanism, the collecting mechanism of hands and the crushing mechanism of bunch stalk. The clamping mechanism for the bunch can be well matched with the transportation ropeway of banana orchards to complete the preparatory work for feeding. The de-handing mechanism realizes the self-adaptive profiling function of the cutting blades to the bunch stalk by means of an iris-like mechanism. The flexible conveyor belt arranged around the center of the de-handing machine can reduce the mechanical damage to the banana hands, and at the same time it can also collect them after de-handing, which is convenient for subsequent centralized cleaning and disinfection. The bunch stalk is dropped on to the crushing mechanism at the bottom of the de-handing machine to complete the cutting and crushing, which is convenient for the subsequent work of returning stalks to the field. The lifting mechanism shown in Figure 12 adopts the design method of combining scissor hinges and screw nut sets, which is convenient for cutting blades to de-hand the banana hands at different positions on the bunch stalk. In addition, the optimally designed radially expandable cutting blade group has higher robustness and stability. In order to make the continuous de-handing process of the whole bunch smoother, Guo et al. designed a de-handing mechanism with expandable function as shown in Figure 13 according to the working principle of the expandable mechanism and with the constant force coil spring [17]. In addition, they also designed a de-handing mechanism that can exert constant clamping force as shown in Figure 14. This feature was achieved using a three-spring parallel mechanism and with the Hooke-like hinge. They are all mechanical de-handing mechanisms with great potential for development at present. All have completed laboratory tests and achieved good results. It is expected that debugging and de-handing work in banana orchards will be carried out as soon as possible.
Based on the comprehensive review of ten mechanical de-handing technologies and equipment discussed above, we scored them in terms of the success rate of de-handing, incision quality and self-adaptive profiling effect (Table 1). Among them, the maximum value of the success rate of de-handing is 100%, and the maximum value of the score of the incision quality and the score of the self-adaptive profiling effect is 10. At the same time, for some technologies and equipment whose experimental results have not been reported, we have also provided expected performance using a banana farmer survey, which can be used as a reference only and needs to be interpreted with caution.
According to the evaluation results, we can conclude that the mechanical equipment with higher potential to achieve a higher success rate in de-handing are shown in Figure 6, Figure 13 and Figure 14; the mechanical equipment with high potential to obtain good incision quality is shown in Figure 6 and Figure 14; the mechanical equipment with high potential to exhibit an excellent self-adaptive profiling performance is shown in Figure 6, Figure 7, Figure 12, Figure 13 and Figure 14. It is worth noting that these evaluation results include two different de-handing methods: the circumferentially rotating mechanical de-handing method and the axial plunge-cutting mechanical de-handing method. At the same time, they also include the de-handing scene for a single banana hand and the continuous de-handing scene of all banana hands in the whole bunch. From these evaluation results, we conclude our survey of the trends and challenges for future research on banana mechanical de-handing technology.

3. Opportunities and Challenges

3.1. Development of Machine Vision Systems

A major advantage of developing machine vision systems is the ability to provide a reliable and efficient automation solution for the banana de-handing operation in a fast and precise manner. As a promising technology to accurately perceive the environment for automated/robotic operation, machine vision has been widely investigated for various applications in fruit crops including phenotyping and canopy detection [19,20,21], branch and trunk identification [22,23,24,25] and fruit localization [26,27]. Different from the industrial environment, fruit orchards are complex, uncertain, variable and with many uncontrollable factors such as light and wind. Lin et al. proposed a fruit detection technology based on contour information in the natural environment [28]. A support vector machine was used as a classifier to detect fruit, which was trained using color and texture features. Aiming at the nighttime orchards working environment, Xiong et al. proposed a method for litchi identification and picking point estimation [29]. The experimental results showed that the accuracy of litchi recognition at night was 93.8% and the average recognition time was 0.52 s. Under different distances to fruit, the accuracy of picking point estimation ranged from 87.5% to 97.5%. Considering the complex environment of orchards and the different light intensities during the day and night, Fu et al. developed a target detection algorithm for automated and rapid detection of kiwifruit in orchards by improving the YOLOv3-tiny model [30,31]. Their research showed that, using the DY3TNet model and along with the YOLOv3-tiny model, better detection performance could be achieved with images captured with artificial lights compared with those without. In addition, Lin et al. proposed a new detection algorithm based on color, depth and shape to detect spherical or cylindrical fruits on plants in natural environment, thereby guiding harvesting robots to automatically pick them [32]. The experimental results showed that, for peppers, eggplants and guava fruits, the detection accuracy was 0.86, 0.87 and 0.89, and the recall was 0.89, 0.76 and 0.81, respectively. In order to better help robots formulate apple picking strategies, Gao et al. proposed a multi-class apple detection method based on Faster Region-Convolutional Neural Network in dense-foliage fruiting-wall trees [33]. This method effectively detected all the different categories of apples and avoided potential damages to branches and lattice lines. In addition, to further improve the apple harvesting system, Zhang et al. developed a multi-class object detection algorithm for automatically detecting apples, branches and tree trunks in natural environment using Faster R-CNN [34]. A total of 72.7% of the vibration locations estimated by the algorithm were deemed appropriate compared to the input from human experts.
At present, there are relatively few reports on research into the visual recognition of banana bunches. Considering the characteristics of low color contrast and complex environment in banana orchards, Chen et al. established a measurement framework based on multi-vision technology and used a set of general methods to improve the comprehensive performance of multi-view geometry-based vision modules in orchard picking tasks [35]. This work provides a theoretical and practical reference for 3D sensing of banana populations in a complex environment. Fu et al. carried out the detection of banana bunches in natural environment based on color and texture features using an RGB camera [36]. The results showed that the average single-scale detection rate based on the proposed algorithm is 89.6%, the average execution time is 1.33 s and the shortest execution time is 0.34 s. In addition, they also proposed a banana bunch detection method based on the latest deep learning algorithm. By using a monocular camera and applying the YOLOv4 neural network algorithm to extract the deep features of the bunch, the accurate detection of bunches of different sizes can be achieved [37]. The algorithm they used achieved a detection accuracy of 99.3% with an average execution time of 0.17 s. Compared with a conventional machine learning algorithm, the deep learning algorithm outperformed both in detection accuracy and time. The experimental results showed that YOLOv4 has higher detection confidence and detection accuracy than YOLOv3. In order to better guide the picking operation of banana bunches in orchards, Wu et al. used an improved deep learning algorithm to identify banana fruits, bunch stalks and flower buds [38]. They used an edge detection algorithm and geometric calculation to determine the picking points on bunch stalk, proposed an improved YOLOv3 model based on clustering optimization and studied the effect of illumination changes on the model. At the same time, Guo et al. conducted discrete element modeling on bunch stalk in order to find out the force changing laws during the picking and continuous de-handing process [39]. Through the method of combining virtual mechanical simulation and physical experiment, the force changing laws of bunch stalk under shear, tension, compression and bending were visualized and the biomechanical properties of bunch stalk were studied. Their research results can provide a reference for probing the mechanism of continuously and mechanically de-handing banana hands.
Rapid identification and precise positioning of the circumferential ring (often irregular) of the fruit node under the complex morphological features of the banana bunch is a key step in the development of machine vision systems for the de-handing operation. Information on the optimal cutting area at the fruit node can be used to guide the cutter during de-handing. The circumferential irregularity of the fruit node is shown in Figure 15. The determination of the optimal cutting area at the fruit node determines in turn the de-handing quality: if the selected cutting area is too close to the bunch stalk, the cutter may cut into the bunch stalk, which reduces the smoothness of the continuous de-handing process. On the other hand, if the selected cutting area is too close to the banana hand, the cutter may cut apart the hand or even cut off the fingers, which is not appropriate for the storage and fresh market sales of the fruit. However, banana bunch stalk, hand and fruit node are all green, and the difference in texture characteristics and color among them is not obvious. It is still challenging to choose an appropriate method to quickly identify them and accurately conduct image segment and extraction of the fruit node. No study, to the best of our knowledge, has been conducted so far on this challenging machine vision task. Barnea et al. proposed a shape-based fruit detective method using RGB and range data to analyze the shape features of objects in the image plane and 3D space [40]. This type of research may offer some help in addressing the challenges in the above questions. Under the complex topography of banana bunches, quickly identifying the fruit node with circumferential irregular ring domain features and establishing a fruit node feature library could be an important area for future research and development. In addition, the high-precision pose solution results and high-reduction 3D reconstruction model are obtained, which is further helpful for guiding it to find and locate the best cutting area on the fruit node.

3.2. Design of the Profiling Cutter

Designing a cutter or end-effector for banana mechanical de-handing is a challenging task because banana hands are often crowded and staggered and the space available for manipulation is very limited. Its design should take into account mechanical and space requirements, such as size, shape, weight and maneuverability; it should also consider physical, horticultural and biological characteristics of plant and fruit. The choice of cutter configuration and power source is critical for developing end-effectors for the de-handing operation. By reviewing the development of mechanical de-handing technology and equipment, it was found that previously developed de-handing cutters still have limitations in terms of kinematic flexibility and topology-optimized design. The end-effector should be compact, lightweight and more flexible. The suitability of the end-effector to integrate with different components such as cameras, cutters and other sensors should also be considered as an important design criterion.
As reported in the literature, the previous designs of de-handing cutters are primarily based on the working principles of compliant mechanisms in the aerospace field or mechanical engineering field, such as the origami [41,42,43,44,45] and expandable mechanism [46,47,48]. These mechanisms usually have the characteristics of compact design and light weight, and when fully unfolded the volume can become several times the initial state. Satellite antennas are the most typical example [49,50,51,52]. Although current industrial manipulators and/or end-effectors have been developed to work in free workspaces, the overall canopy of the banana bunch is unstructured and complex, and there are narrow spaces between banana hands to maneuver, resulting in many challenges in developing cutters or end-effectors that can function in narrow space. In addition, the cutting end-effector is more likely to collide with the bunch stalk and fruit due to the proximity of the de-handing area to bunch stalk and banana fingers. This may damage both the de-handing machine and banana hands, reducing the quality of fruit and being unfavorable for storage and fresh market sales. It is also noted that, by observing the morphological characteristics of the bunch and studying the mechanical de-handing methods, we found that, although the structures of the previous de-handing cutters were not optimal, the design ideas were innovative and applicable. We believe that future research could focus on obtaining the circumferentially expandable cutter and the discrete element mechanical model of the irregular ring of fruit node by integrating the method of mechanism topology design optimization and the discrete element modeling method of biomechanical characteristics of the plant, and then research the cutter mechanism in cutting behavior of the banana fruit node.

3.3. Robotic De-Handing

With the rapid advancement in robotic, computational and machine learning techniques in recent years, robotic operations in fruit orchards with eye-and-hand co-ordination to identify, locate and pick fruits have been a promising area of research and development and eventually commercial adoption [53,54,55,56,57]. Major steps in robotic orchard operations include: (i) real-time identification of fruits and other canopy parts in orchards using a machine vision or perception system and estimation of their precise spatial coordinates; (ii) path planning to move the manipulator such that the end-effector is guided to the designated work area to complete the corresponding task while avoiding obstacles; and (iii) performing the specific field operation. For example, detection of apples in dense branches [58], the routing of guava picking robots in orchards [59], rapid identification of litchi clusters on branches [60] and precise picking of kiwifruit [61] are a few studies including these steps in robotic harvesting of fruit. However, the research and application of such a robotic system in banana de-handing operation has not yet been reported.
The positioning of the optimal cutting area of the fruit node based on visual perception and cutter coupling is critical to the success of mechanical de-handing. Because it guides the cutter to cut the fruit node accurately, a more precise position and a minimal positioning error are required to avoid cutting into bunch stalk or cutting banana hands apart. Efficient identification of obstacles is a prerequisite for path planning, and the information on the cutting area will be used by the mechanical system to perform the de-handing task. Sensing systems and machine vision systems were used to collect information (recognition and localization) in the complex topographic environment of the bunch, and then path planning and control schemes were used to manipulate mechanical systems (robot arms or end effectors) to reach the targeted cutting area. The locating of the optimal de-handing area on fruit node is still a challenging task, which makes the eye-hand synergistic de-handing system still at the development stage. Overall, AI decides where to de-hand the banana hand and the end-effector reaches the designated location on the fruit node to cut, which is our intended goal. This research direction of robotic de-handing is very challenging, and no studies have been reported in the past, as mentioned previously. Therefore, future studies could focus on real-time path planning of the cutter, closed-loop feedback control with visual serving and error compensation of the spatial dynamic pose of the expandable de-handing cutter. In addition, studies could be conducted to obtain the optimal cutting position on the fruit node accurately, and finally to investigate the coupled de-handing mechanism of visual locating and profiling cutting of the banana fruit node.

3.4. Fruit Quality Assessment

In the process of banana de-handing, banana hands may be impacted by mechanical devices to varying degrees, resulting in mechanical damage. In addition, mis-operations such as the irregular cutting behavior of the de-handing cutter on the fruit node may cause potential mechanical damage to banana fingers. With the exception of fruit quality issues caused by bruising banana fingers, cutting banana hands apart and cutting fingers into pieces, some areas of mechanical damage were too small to be detected by the human eye. Therefore, in the process of mechanical de-handing, in addition to reducing the mechanical damage of banana hands as much as possible, it is also essential to evaluate the mechanical damage of banana hands after de-handing. This process can help banana farmers to select and grade the fruits, and arrange storage and sales in a timely fashion.
Spectral imaging technology, as a rapid, non-destructive and non-contact detective method, has been widely used in farmland water/nutrient detection [62,63,64,65,66], crop quality analysis [67,68,69,70] and crop phenotype research [71], among others. In terms of fruit quality assessment, Xu et al. proposed a method for fruit mechanical damage recognition based on the fusion of hyperspectral sensing and electronic nose [72]. The research results showed that the accuracy of this method for damage detection in guava was 97.4%, which showed the feasibility of multi-source information fusion for improving the performance of fruit mechanical damage detection. This type of detection method may be applied to detect the mechanical damage of banana hands after de-handing. Aiming at the mechanical damage of litchi caused by impact during vibration harvesting and transportation, researchers have established an elastic impact model and a model that considers energy dissipation to describe the deformation behavior of litchi during the impact process. By studying the behavior of fruit-to-fruit and fruit-to-rigid plate collisions, a method to evaluate the impact tolerance and extent of fruit damage was obtained. In addition, the shedding law and influence characteristics of litchi fruits under the vibration harvesting method were studied [73,74,75,76].
Another area of research that can support analysis of mechanical impact and its damage include the finite element [77,78] or discrete element modeling of the fruit [79,80]. This technique can visualize stress developed in the fruit surface due to impacts, and the damage of the fruit can be predicted and evaluated by means of simulation analysis. Although there is no report on the prediction and evaluation of banana mechanical damage after de-handing, non-destructive detecting technology and simulation analysis methods are effective means with long-term availability and potential [81]. They can be a good help for researchers or banana farmers to assess the damage to banana hand after de-handing, and provide them with reasonable and effective treatment suggestions.
In summary, it is considered that the road to practically adopting mechanical or robotic de-handing techniques in banana plantation is long and full of challenges. Focused research and development must be continued in a strategic manner to advance the engineering solutions, while also assessing economic and practical viability, in challenging banana orchard environments [82]. This paper has proposed some crucial directions for future research and development in banana mechanical de-handing, which is expected to bring the technology further towards potential commercialization.

4. Summary and Conclusions

Mechanical de-handing is a potential long-term solution to the existing labor shortage; however, the research on mechanical de-handing technology and the development of mechanical equipment are facing challenges. This paper systematically reviewed the state-of-the-art technologies investigated for mechanized banana de-handing, which fell into one of two categories: (i) circumferentially rotating mechanical de-handing, or (ii) axial plunge-cutting mechanical de-handing, selected from the current literature. These techniques were assessed in terms of the success rate of de-handing, incision quality and self-adaptive profiling effect, which served as the criteria for defining their performance. This review found four key challenges in current de-handing technologies, which also provide opportunities for further research and development: (i) autonomous/automated visual perception; (ii) reliable profile cutting; (iii) integrated robotic system with eye–hand synergy/coordination, and (iv) fruit quality assessment after mechanical de-handing.
Mechanical de-handing is challenging due to the complex morphology and structure of banana bunches. Critical survey of the available literature suggested that the current research has not fully addressed the challenges in implementing mechanical de-handing equipment in banana postharvest operations, which revealed the hurdles that need to be overcome for practical adoption of this technology in banana orchards. In this way, these challenges are suggested as directions for future research in mechanical de-handing. Development of sensing/vision, computational and robotic technology, mechanical/automated de-handing or even smart de-handing techniques may lead to commercial viability of this technique, which is expected to contribute to the sustainable development of the banana industry. It is also noted that, for robust and reliable system development, an integrated, trans-disciplinary approach needs to be considered including expertise from agricultural engineering, agronomy/horticulture, computer science and economics.

Author Contributions

J.G.: Conceptualization, Investigation, Methodology, Writing—original draft, Writing—review & editing. J.D.: Funding acquisition, Project administration, Resources, Writing—review & editing. Z.Y.: Funding acquisition, Project administration, Resources, Writing—review & editing. M.K.: Writing—review & editing. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Laboratory of Lingnan Modern Agriculture Project (NT2021009), the National Key Research and Development Program of China (2020YFD1000104), the China Agriculture Research System of MOF and MARA (CARS-31) and Guangdong Provincial Special Fund for Modern Agriculture Industry Technology Innovation Teams (2022KJ109). The China Scholarship Council (CSC) sponsored Jie Guo (File No. 202008440665) in conducting his collaborative doctoral dissertation studies at the WSU Center for Precision and Automated Agricultural Systems (CPAAS). The authors wish to thank their generous financial assistance.

Institutional Review Board Statement

Not applicable.

Data Availability Statement

Not applicable.

Acknowledgments

The authors thank the editor and reviewers for their valuable input that helped improve the quality of the manuscript. We also thank Xiangjun Zou, Jun Li and Han Fu who offered valuable feedback on the content of the article and thus contributed to its quality.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Huang, Y.; Xu, X. Current situation and development trend of banana industry in global. Trop. Agric. Eng. 2021, 45, 34–38. [Google Scholar]
  2. Ke, Y.; Guo, J.; Fang, J.; Ma, C.; Zhang, H. Analysis on the development trend of world banana production and trade. World Agric. 2008, 4, 38–39, 67. [Google Scholar]
  3. Xie, J. Fruit scientific research in New China in the past 70 years: Banana. J. Fruit Sci. 2019, 36, 1429–1440. [Google Scholar] [CrossRef]
  4. Guo, J.; Duan, J.; Li, J.; Yang, Z. Mechanized technology research and equipment application of banana post-harvesting: A review. Agronomy 2020, 10, 374. [Google Scholar] [CrossRef] [Green Version]
  5. Luo, X.; Liao, J.; Hu, L.; Zang, Y.; Zhou, Z. Improving agricultural mechanization level to promote agricultural sustainable development. Trans. Chin. Soc. Agric. Eng. 2016, 32, 1–11. [Google Scholar] [CrossRef]
  6. Guo, J.; Fu, H.; Yang, Z.; Li, J.; Jiang, Y.; Jiang, T.; Liu, E.; Duan, J. Research on the physical characteristic parameters of banana bunches for the design and development of post-harvesting machinery and equipment. Agriculture 2021, 11, 362. [Google Scholar] [CrossRef]
  7. Merino, G.; da Silva, L.; Mattos, D.; Guimarães, B.; Merino, E. Ergonomic evaluation of the musculoskeletal risks in a banana harvesting activity through qualitative and quantitative measures, with emphasis on motion capture (Xsens) and EMG. Int. J. Ind. Ergon. 2019, 69, 80–89. [Google Scholar] [CrossRef]
  8. Guo, J.; Karkee, M.; Yang, Z.; Fu, H.; Li, J.; Jiang, Y.; Jiang, T.; Liu, E.; Duan, J. Research of simulation analysis and experimental optimization of banana de-handing device with self-adaptive profiling function. Comput. Electron. Agric. 2021, 185, 106148. [Google Scholar] [CrossRef]
  9. Yang, Z.; Guo, J.; Jin, M.; Xu, Z.; Duan, J.; Mo, J.; Jin, H. A variable diameter self-adaptive rotary cutting banana de-handing cutter. China Pat. 2019, 11, 1–8. [Google Scholar]
  10. Xu, Z.; Yang, Z.; Duan, J.; Jin, M.; Mo, J.; Zhao, L.; Guo, J.; Yao, H. Design and experiment of symmetrical shape deployable arc profiling mechanism based on composite multi-cam structure. Symmetry 2019, 11, 958. [Google Scholar] [CrossRef] [Green Version]
  11. Duan, J.; Guo, J.; Yang, Z.; Fu, H.; L, J.; Jin, M.; Zhao, L.; Jiang, Y.; Jiang, T.; Ding, Y.; et al. A de-handing platform for self-adaptive banana bunch stalk. China Pat. 2021, 9, 1–9. [Google Scholar]
  12. Yang, Z.; Guo, J.; Jin, M.; Xu, Z.; Duan, J.; Mo, J.; Jin, H. A plunge-cut banana de-handing mechanism that with self-adaptive function to embrace bunch stalk. China Pat. 2019, 8, 1–8. [Google Scholar]
  13. Yang, Z.; Guo, J.; Jin, M.; Xu, Z.; Duan, J.; Mo, J.; Jin, H. A variable-diameter plunge-cut banana de-handing mechanism. China Pat. 2020, 12, 1–8. [Google Scholar]
  14. Yang, Z.; Guo, J.; Duan, J.; Jin, M.; Fu, H.; Sun, J.; Xie, H.; Mo, J.; Zhao, L.; Jiang, T. A variable diameter self-adaptive banana bunch stalk de-handing cutter based on wire cutting. China Pat. 2021, 2, 1–8. [Google Scholar]
  15. Yang, Z.; Guo, J.; Fu, H.; Jin, M.; Duan, J.; Sun, J.; Xie, H.; Zhao, L.; Xu, Z.; Ding, Y.; et al. Variable diameter plunge-cutting banana de-handing machine based on iris-gear-like mechanism. China Pat. 2020, 7, 1–13. [Google Scholar]
  16. Yang, Z.; Guo, J.; Duan, J.; Fu, H.; Li, J.; Jin, M.; Zhao, L.; Jiang, Y.; Jiang, T.; Liu, E.; et al. A synchronous self-centering and lifting straight-line cutting banana de-handing mechanism. China Pat. 2021, 12, 1–10. [Google Scholar]
  17. Guo, J.; Jin, M.; Duan, J.; Li, J.; Fu, H.; Zhao, L.; Xu, Z.; Ding, Y.; Jiang, Y.; Yang, Z. Design and experiment of banana de-handing device based on symmetrical shape deployable mechanism. Symmetry 2020, 12, 415. [Google Scholar] [CrossRef] [Green Version]
  18. Yang, Z.; Guo, J.; Jin, M.; Duan, J.; Fu, H.; Xu, Z. Design and analysis of self-adaptability in banana crown-cutting device based on constant-force mechanism. Trans. Chin. Soc. Agric. Mach. 2019, 50, 148–155. [Google Scholar] [CrossRef]
  19. Chen, M.; Tang, Y.; Zou, X.; Huang, Z.; Zhou, H.; Chen, S. 3D global mapping of large-scale unstructured orchard integrating eye-in-hand stereo vision and SLAM. Comput. Electron. Agric. 2021, 187, 106237. [Google Scholar] [CrossRef]
  20. Huang, Y.; Ren, Z.; Li, D.; Liu, X. Phenotypic techniques and applications in fruit trees: A review. Plant Methods 2020, 16, 1–22. [Google Scholar] [CrossRef]
  21. Zhang, X.; He, L.; Zhang, J.; Whiting, M.D.; Karkee, M.; Zhang, Q. Determination of key canopy parameters for mass mechanical apple harvesting using supervised machine learning and principal component analysis (PCA). Biosyst. Eng. 2020, 193, 247–263. [Google Scholar] [CrossRef]
  22. Lin, G.; Tang, Y.; Zou, X.; Wang, C. Three-dimensional reconstruction of guava fruits and branches using instance segmentation and geometry analysis. Comput. Electron. Agric. 2021, 184, 106107. [Google Scholar] [CrossRef]
  23. Zhang, X.; Karkee, M.; Zhang, Q.; Whiting, M.D. Computer vision-based tree trunk and branch identification and shaking points detection in Dense-Foliage canopy for automated harvesting of apples. J. Field Robot. 2021, 38, 476–493. [Google Scholar] [CrossRef]
  24. Li, J.; Tang, Y.; Zou, X.; Lin, G.; Wang, H. Detection of fruit-bearing branches and localization of litchi clusters for vision-based harvesting robots. IEEE Access 2020, 8, 117746–117758. [Google Scholar] [CrossRef]
  25. Majeed, Y.; Karkee, M.; Zhang, Q. Estimating the trajectories of vine cordons in full foliage canopies for automated green shoot thinning in vineyards. Comput. Electron. Agric. 2020, 176, 105671. [Google Scholar] [CrossRef]
  26. Fu, L.; Gao, F.; Wu, J.; Li, R.; Karkee, M.; Zhang, Q. Application of consumer RGB-D cameras for fruit detection and localization in field: A critical review. Comput. Electron. Agric. 2020, 177, 105687. [Google Scholar] [CrossRef]
  27. Tang, Y.; Chen, M.; Wang, C.; Luo, L.; Li, J.; Lian, G.; Zou, X. Recognition and localization methods for vision-based fruit picking robots: A Review. Front. Plant Sci. 2020, 11, 510. [Google Scholar] [CrossRef]
  28. Lin, G.; Tang, Y.; Zou, X.; Cheng, J.; Xiong, J. Fruit detection in natural environment using partial shape matching and probabilistic hough transform. Precis. Agric. 2020, 21, 160–177. [Google Scholar] [CrossRef]
  29. Xiong, J.; Lin, R.; Liu, Z.; He, Z.; Tang, L.; Yang, Z.; Zou, X. The recognition of litchi clusters and the calculation of picking point in a nocturnal natural environment. Biosyst. Eng. 2018, 166, 44. [Google Scholar] [CrossRef]
  30. Fu, L.; Feng, Y.; Wu, J.; Liu, Z.; Gao, F.; Majeed, Y.; Al-Mallahi, A.; Zhang, Q.; Li, R.; Cui, Y. Fast and accurate detection of kiwifruit in orchard using improved YOLOv3-tiny model. Precis. Agric. 2021, 22, 754–776. [Google Scholar] [CrossRef]
  31. Fu, L.; Tola, E.; Al-Mallahi, A.; Li, R.; Cui, Y. A novel image processing algorithm to separate linearly clustered kiwifruits. Biosyst. Eng. 2019, 183, 184–195. [Google Scholar] [CrossRef]
  32. Lin, G.; Tang, Y.; Zou, X.; Xiong, J.; Fang, Y. Color-, depth-, and shape-based 3D fruit detection. Precis. Agric. 2020, 21, 1–17. [Google Scholar] [CrossRef]
  33. Gao, F.; Fu, L.; Zhang, X.; Majeed, Y.; Li, R.; Karkee, M.; Zhang, Q. Multi-class fruit-on-plant detection for apple in SNAP system using Faster R-CNN. Comput. Electron. Agric. 2020, 176, 105634. [Google Scholar] [CrossRef]
  34. Zhang, J.; Karkee, M.; Zhang, Q.; Zhang, X.; Yaqoob, M.; Fu, L.; Wang, S. Multi-class object detection using faster R-CNN and estimation of shaking locations for automated shake-and-catch apple harvesting. Comput. Electron. Agric. 2020, 173, 105384. [Google Scholar] [CrossRef]
  35. Chen, M.; Tang, Y.; Zou, X.; Huang, K.; Huang, Z.; Zhou, H.; Wang, C.; Lian, G. Three-dimensional perception of orchard banana central stock enhanced by adaptive multi-vision technology. Comput. Electron. Agric. 2020, 174, 105508. [Google Scholar] [CrossRef]
  36. Fu, L.; Duan, J.; Zou, X.; Lin, G.; Song, S.; Ji, B.; Yang, Z. Banana detection based on color and texture features in the natural environment. Comput. Electron. Agric. 2019, 167, 105057. [Google Scholar] [CrossRef]
  37. Fu, L.; Duan, J.; Zou, X.; Lin, J.; Zhao, L.; Li, J.; Yang, Z. Fast and accurate detection of banana fruits in complex background orchards. IEEE Access 2020, 8, 196835–196846. [Google Scholar] [CrossRef]
  38. Wu, F.; Duan, J.; Chen, S.; Ye, Y.; Ai, P.; Yang, Z. Multi-target recognition of bananas and automatic positioning for the inflorescence axis cutting point. Front. Plant Sci. 2021, 12, 705021. [Google Scholar] [CrossRef]
  39. Guo, J.; Karkee, M.; Yang, Z.; Fu, H.; Li, J.; Jiang, Y.; Jiang, T.; Liu, E.; Duan, J. Discrete element modeling and physical experiment research on the biomechanical properties of banana bunch stalk for postharvest machine development. Comput. Electron. Agric. 2021, 188, 106308. [Google Scholar] [CrossRef]
  40. Barnea, E.; Mairon, R.; Ben-Shahar, O. Colour-agnostic shape-based 3D fruit detection for crop harvesting robots. Biosyst. Eng. 2016, 146, 57–70. [Google Scholar] [CrossRef]
  41. Chen, Y.; Yan, J.; Feng, J. Geometric and kinematic analyses and novel characteristics of origami-inspired structures. Symmetry 2019, 11, 1101. [Google Scholar] [CrossRef] [Green Version]
  42. Lee, D.Y.; Kim, S.R.; Kim, J.S.; Park, J.J.; Cho, K.J. Origami wheel transformer: A variable-diameter wheel drive robot using an origami structure. Soft Robot. 2017, 4, 163–180. [Google Scholar] [CrossRef]
  43. Ngo, T.H.; Chi, I.T.; Chau, M.Q.; Wang, D.A. An energy harvester based on a bistable origami mechanism. Int. J. Precis. Eng. Man. 2022, 23, 213–226. [Google Scholar] [CrossRef]
  44. Nathan, C.B.; Collin, Y.; Amanda, L.; Larry, L.H.; Spencer, P.M. Approaches for minimizing joints in single-degree-of-freedom origami-based mechanisms. J. Mech. Des. 2022, 144, 103301. [Google Scholar] [CrossRef]
  45. Larissa, M.F.; Guilherme, V.R.; Marcelo, A.S. An overview of the mechanical description of origami-inspired systems and structures. Int. J. Mech. Sci. 2022, 223, 107316. [Google Scholar] [CrossRef]
  46. Huang, H.; Li, B.; Zhang, T.; Zhang, Z.; Qi, X.; Hu, Y. Design of large single-mobility surface-deployable mechanism using irregularly shaped triangular prismoid modules. J. Mech. Des. Trans. ASME 2019, 141, 012301. [Google Scholar] [CrossRef]
  47. Cao, W.A.; Yang, D.; Ding, H. Topological structural design of umbrella-shaped deployable mechanisms based on new spatial closed-loop linkage units. J. Mech. Des. Trans. ASME 2018, 140, 062302. [Google Scholar] [CrossRef]
  48. Qi, X.; Huang, H.; Miao, Z.; Li, B.; Deng, Z. Design and mobility analysis of large deployable mechanisms based on plane-symmetric bricard linkage. J. Mech. Des. Trans. ASME 2017, 139, 022302. [Google Scholar] [CrossRef]
  49. Han, B.; Xu, Y.; Yao, J.; Zheng, D.; Guo, X.; Zhao, Y. Configuration synthesis of hoop truss deployable mechanisms for space antenna based on screw theory. AIP Adv. 2019, 9, 085201. [Google Scholar] [CrossRef]
  50. Han, B.; Xu, Y.; Yao, J.; Zheng, D.; Li, Y.; Zhao, Y. Design and analysis of a scissors double-ring truss deployable mechanism for space antennas. Aerosp. Sci. Technol. 2019, 93, 105357. [Google Scholar] [CrossRef]
  51. Qi, X.; Huang, H.; Li, B.; Deng, Z. A large ring deployable mechanism for space satellite antenna. Aerosp. Sci. Technol. 2016, 58, 498–510. [Google Scholar] [CrossRef]
  52. Xu, Y.; Guan, F.; Xu, X.; Wang, H.; Zheng, Y. Development of a novel double-ring deployable mesh antenna. Int. J. Antennas Propag. 2012, 11, 1497–1500. [Google Scholar] [CrossRef] [Green Version]
  53. Humna, K.; Travis, J.E.; Aitazaz, A.F.; Farhat, A. Wild blueberry harvesting losses predicted with selective machine learning algorithms. Agriculture 2022, 12, 1657. [Google Scholar] [CrossRef]
  54. Bayu, T.P.; Krishna, S.I.; Buyung, A.F. Development of a handheld iot-based fruit harvester to support agrotourism. Microprocess. Microsy. 2022, 91, 104550. [Google Scholar] [CrossRef]
  55. Denchai, W.; Panarit, S.; Surasak, C. Automatic classification of the ripeness stage of mango fruit using a machine learning approach. Agri. Eng. 2022, 4, 32–47. [Google Scholar] [CrossRef]
  56. Meenakshi, S.K.; Santhakumar, M. Selective fruit harvesting: Research, trends and developments towards fruit detection and localization–A review. P. I Mech. Eng. C-J. Mec. 2022, 10, 09544062221128443. [Google Scholar] [CrossRef]
  57. Zhou, H.; Xiao, J.; Kang, H.; Wang, X.; Au, W.; Chen, C. Learning-based slip detection for robotic fruit grasping and manipulation under leaf interference. Sensors 2022, 22, 5483. [Google Scholar] [CrossRef]
  58. Fu, L.; Majeed, Y.; Zhang, X.; Karkee, M.; Zhang, Q. Faster R–CNN–based apple detection in dense-foliage fruiting-wall trees using RGB and depth features for robotic harvesting. Biosyst. Eng. 2020, 197, 245–256. [Google Scholar] [CrossRef]
  59. Lin, G.; Zhu, L.; Li, J.; Zou, X.; Tang, Y. Collision-free path planning for a guava-harvesting robot based on recurrent deep reinforcement learning. Comput. Electron. Agric. 2021, 188, 106350. [Google Scholar] [CrossRef]
  60. Ye, L.; Duan, J.; Yang, Z.; Zou, X.; Chen, M.; Zhang, S. Collision-free motion planning for the litchi-picking robot. Comput. Electron. Agric. 2021, 185, 106151. [Google Scholar] [CrossRef]
  61. Williams, H.A.M.; Jones, M.H.; Nejati, M.; Seabright, M.J.; Bell, J.; Penhall, N.D.; Barnett, J.J.; Duke, M.D.; Scarfe, A.J.; Ahn, H.S.; et al. Robotic kiwifruit harvesting using machine vision, convolutional neural networks, and robotic arms. Biosyst. Eng. 2019, 181, 140–156. [Google Scholar] [CrossRef]
  62. Wang, C.; Liu, B.; Liu, L.; Zhu, Y.; Hou, J.; Liu, P.; Li, X. A review of deep learning used in the hyperspectral image analysis for agriculture. Artif. Intell. Rev. 2021, 54, 5205–5253. [Google Scholar] [CrossRef]
  63. Lu, B.; Dao, P.D.; Liu, J.; He, Y.; Shang, J. Recent advances of hyperspectral imaging technology and applications in agriculture. Remote Sens. 2020, 12, 2659. [Google Scholar] [CrossRef]
  64. Grieco, M.; Schmidt, M.; Warnemünde, S.; Backhaus, A.; Klück, H.C.; Garibay, A.; Moya, Y.A.; Jozefowicz, A.M.; Mock, H.P.; Seiffert, U.; et al. Dynamics and genetic regulation of leaf nutrient concentration in barley based on hyperspectral imaging and machine learning. Plant Sci. 2022, 315, 111123. [Google Scholar] [CrossRef] [PubMed]
  65. Guo, Y.; Chen, S.; Li, X.; Cunha, M.; Jayavelu, S.; Cammarano, D.; Fu, Y. Machine learning-based approaches for predicting SPAD values of maize using multi-spectral images. Remote Sens. 2022, 14, 1337. [Google Scholar] [CrossRef]
  66. Chang, S.; Lee, U.; Kim, J.B.; Jo, Y.D. Application of 3D-volumetric analysis and hyperspectral imaging systems for investigation of heterosis and cytoplasmic effects in pepper. Sci. Hortic. 2022, 302, 111150. [Google Scholar] [CrossRef]
  67. Wang, F.; Zhao, C.; Xu, B.; Xu, Z.; Li, Z.; Yang, H.; Duan, D.; Yang, G. Development of a portable detection device for the quality of fresh tea leaves using spectral technology. Trans. Chin. Soc. Agric. Eng. 2020, 36, 273–280. [Google Scholar] [CrossRef]
  68. Fu, X.; Ying, Y. Application of NIR and Raman spectroscopy for quality and safety inspection of fruits and vegetables: A Review. Trans. Chin. Soc. Agric. Mach. 2013, 44, 148–164. [Google Scholar] [CrossRef]
  69. Ahmad, U.; Nasirahmadi, A.; Hensel, O.; Marino, S. Technology and data fusion methods to enhance site-specific crop monitoring. Agronomy 2022, 12, 555. [Google Scholar] [CrossRef]
  70. Li, Y.; Chang, C.; Wang, Z.; Li, T.; Li, J.; Zhao, G. Identification of cultivated land quality grade using fused multi-source data and multi-temporal crop remote sensing information. Remote Sens. 2022, 14, 2109. [Google Scholar] [CrossRef]
  71. Sankaran, S.; Khot, L.R.; Espinoza, C.Z.; Jarolmasjed, S.; Sathuvalli, V.R.; Vandemark, G.J.; Miklas, P.N.; Carter, A.H.; Pumphrey, M.O.; Knowles, R.R.N.; et al. Low-altitude, high-resolution aerial imaging systems for row and field crop phenotyping: A review. Eur. J. Agron. 2015, 70, 112–123. [Google Scholar] [CrossRef]
  72. Xu, S.; Lu, H.; Zhou, Z.; Lv, E.; Jiang, Y. Identification for guava mechanical damage based on combined hyper spectrometer and electronic nose. Trans. Chin. Soc. Agric. Mach. 2015, 46, 214–219. [Google Scholar] [CrossRef]
  73. Wang, Y.; Wang, W.; Fu, H.; Yang, Z.; Lu, H. Detachment patterns and impact characteristics of litchi fruit during vibrational harvesting. Sci. Hortic. Amst. 2022, 295, 110836. [Google Scholar] [CrossRef]
  74. Wang, W.; Zhang, S.; Fu, H.; Lu, H.; Yang, Z. Evaluation of litchi impact damage degree and damage susceptibility. Comput. Electron. Agric. 2020, 173, 105409. [Google Scholar] [CrossRef]
  75. Wang, W.; Lu, H.; Zhang, S.; Yang, Z. Damage caused by multiple impacts of litchi fruits during vibration harvesting. Comput. Electron. Agric. 2019, 162, 732–738. [Google Scholar] [CrossRef]
  76. Zhang, S.; Wang, W.; Wang, Y.; Fu, H.; Yang, Z. Improved prediction of litchi impact characteristics with an energy dissipation model. Postharvest Biol. Technol. 2021, 176, 111508. [Google Scholar] [CrossRef]
  77. Celik, H.K. Determination of bruise susceptibility of pears (Ankara variety) to impact load by means of FEM-based explicit dynamics simulation. Postharvest Biol. Technol. 2017, 128, 83–97. [Google Scholar] [CrossRef]
  78. Li, Z.; Wang, Y. A multiscale finite element model for mechanical response of tomato fruits. Postharvest Biol. Technol. 2016, 121, 19–26. [Google Scholar] [CrossRef]
  79. Scheffler, O.C.; Coetzee, C.J.; Opara, U.L. A discrete element model (DEM) for predicting apple damage during handling. Biosyst. Eng. 2018, 172, 29–48. [Google Scholar] [CrossRef]
  80. Li, X.; Du, Y.; Liu, L.; Mao, E.; Yang, F.; Wu, J.; Wang, L. Research on the constitutive model of low-damage corn threshing based on DEM. Comput. Electron. Agric. 2022, 194, 106722. [Google Scholar] [CrossRef]
  81. Fu, H.; Du, W.; Yang, J.; Wang, W.; Wu, Z.; Yang, Z. Bruise measurement of fresh market apples caused by repeated impacts using a pendulum method. Postharvest Biol. Technol. 2022, 195, 112143. [Google Scholar] [CrossRef]
  82. Luo, X.; Liao, J.; Zou, X.; Zhang, Z.; Zhou, Z.; Zang, Y.; Hu, L. Enhancing agricultural mechanization level through information technology. Trans. Chin. Soc. Agric. Eng. 2016, 32, 1–14. [Google Scholar] [CrossRef]
Agriculture 12 01821 g001 550
Figure 1. A general flow of banana harvest and postharvest operations.
Figure 1. A general flow of banana harvest and postharvest operations.
Agriculture 12 01821 g001
Agriculture 12 01821 g002 550
Figure 2. Banana orchards and their growing patterns. (a) Appearance of a banana orchard located in a hilly mountainous area, (b) The interior view of a banana plantation in the plains.
Figure 2. Banana orchards and their growing patterns. (a) Appearance of a banana orchard located in a hilly mountainous area, (b) The interior view of a banana plantation in the plains.
Agriculture 12 01821 g002
Agriculture 12 01821 g003 550
Figure 3. Manual de-handing mode of banana bunches using a simple cutting tool.
Figure 3. Manual de-handing mode of banana bunches using a simple cutting tool.
Agriculture 12 01821 g003
Agriculture 12 01821 g004 550
Figure 4. Schematic diagram of manual de-handing [7] (a); and virtual prototype of mechanical de-handing systems [8] (b).
Figure 4. Schematic diagram of manual de-handing [7] (a); and virtual prototype of mechanical de-handing systems [8] (b).
Agriculture 12 01821 g004
Agriculture 12 01821 g005 550
Figure 5. Rotary de-handing cutter with variable enveloping diameter [9]. (a) 3D model, (b) Front view, (c) Bottom view, (d) Side view.
Figure 5. Rotary de-handing cutter with variable enveloping diameter [9]. (a) 3D model, (b) Front view, (c) Bottom view, (d) Side view.
Agriculture 12 01821 g005
Agriculture 12 01821 g006 550
Figure 6. Design and application of the circumferentially expandable de-handing platform. (a) Schematic diagram of the profiling mechanism, (b) Prototype of the de-handing platform, (c) Schematic of an indoor experiment, (d) Incision of banana hands with different profiling radius.
Figure 6. Design and application of the circumferentially expandable de-handing platform. (a) Schematic diagram of the profiling mechanism, (b) Prototype of the de-handing platform, (c) Schematic of an indoor experiment, (d) Incision of banana hands with different profiling radius.
Agriculture 12 01821 g006
Agriculture 12 01821 g007 550
Figure 7. Schematic diagram of de-handing platform that can adapt to variable diameters of bunch stalk. (a) Virtual prototype, (b) Top view, (c) Left view, (d) Axonometric view of key components.
Figure 7. Schematic diagram of de-handing platform that can adapt to variable diameters of bunch stalk. (a) Virtual prototype, (b) Top view, (c) Left view, (d) Axonometric view of key components.
Agriculture 12 01821 g007
Agriculture 12 01821 g008 550
Figure 8. A de-handing mechanism that can adapt to changes in the diameter of bunch stalk. (a) Axonometric view, (b) Front view, (c) Top view, (d) Axonometric view of key components.
Figure 8. A de-handing mechanism that can adapt to changes in the diameter of bunch stalk. (a) Axonometric view, (b) Front view, (c) Top view, (d) Axonometric view of key components.
Agriculture 12 01821 g008
Agriculture 12 01821 g009 550
Figure 9. Plunge-cutting de-handing mechanism with variable diameter. (a) Side view, (b) Front view, (c) Top view, (d) Side view of key components.
Figure 9. Plunge-cutting de-handing mechanism with variable diameter. (a) Side view, (b) Front view, (c) Top view, (d) Side view of key components.
Agriculture 12 01821 g009
Agriculture 12 01821 g010 550
Figure 10. De-handing cutter based on wire-cutting. (a) Axonometric view, (b) Assembly schematic of the wire, coil spring and ball, (c) Front view, (d) Structure diagram of the casing of the retractor.
Figure 10. De-handing cutter based on wire-cutting. (a) Axonometric view, (b) Assembly schematic of the wire, coil spring and ball, (c) Front view, (d) Structure diagram of the casing of the retractor.
Agriculture 12 01821 g010
Agriculture 12 01821 g011 550
Figure 11. De-handing machine based on iris-gear-like mechanism. (a) Schematic diagram of the structure of de-handing machine, (b) Axonometric view of de-handing mechanism, (c) Top view of de-handing mechanism, (d) 3D model of arc blade.
Figure 11. De-handing machine based on iris-gear-like mechanism. (a) Schematic diagram of the structure of de-handing machine, (b) Axonometric view of de-handing mechanism, (c) Top view of de-handing mechanism, (d) 3D model of arc blade.
Agriculture 12 01821 g011
Agriculture 12 01821 g012 550
Figure 12. The de-handing mechanism with the functions of radial synchronous movement of blades and vertical lifting of the axial direction. (a) Schematic diagram of the structure of de-handing mechanism, (b) Front view of de-handing mechanism, (c) Top view of de-handing mechanism, (d) Partial enlarged view of the structure of the lifting mechanism.
Figure 12. The de-handing mechanism with the functions of radial synchronous movement of blades and vertical lifting of the axial direction. (a) Schematic diagram of the structure of de-handing mechanism, (b) Front view of de-handing mechanism, (c) Top view of de-handing mechanism, (d) Partial enlarged view of the structure of the lifting mechanism.
Agriculture 12 01821 g012
Agriculture 12 01821 g013 550
Figure 13. De-handing mechanism based on expandable mechanism. (a) Axonometric view of de-handing mechanism, (b) Front view of de-handing mechanism, (c) Radial expanding process of cutting blade group, (d) Schematic diagram of de-handing experiment, (e) Incision of banana hands after de-handing.
Figure 13. De-handing mechanism based on expandable mechanism. (a) Axonometric view of de-handing mechanism, (b) Front view of de-handing mechanism, (c) Radial expanding process of cutting blade group, (d) Schematic diagram of de-handing experiment, (e) Incision of banana hands after de-handing.
Agriculture 12 01821 g013
Agriculture 12 01821 g014 550
Figure 14. De-handing mechanism based on constant force mechanism. (a) Axonometric view of de-handing mechanism, (b) Top view of de-handing mechanism, (c) Schematic diagram of de-handing experiment, (d) Incision of banana hands after de-handing.
Figure 14. De-handing mechanism based on constant force mechanism. (a) Axonometric view of de-handing mechanism, (b) Top view of de-handing mechanism, (c) Schematic diagram of de-handing experiment, (d) Incision of banana hands after de-handing.
Agriculture 12 01821 g014
Agriculture 12 01821 g015 550
Figure 15. Irregular ring domains at the fruit node. (a) The appearance of the fruit node when looking up, (b) The appearance of the fruit node in bird’s-eye view.
Figure 15. Irregular ring domains at the fruit node. (a) The appearance of the fruit node when looking up, (b) The appearance of the fruit node in bird’s-eye view.
Agriculture 12 01821 g015
Table
Table 1. Performance measures for various mechanical de-handing technology and equipment reviewed in this article.
Table 1. Performance measures for various mechanical de-handing technology and equipment reviewed in this article.
De-Handing MethodFigure NumberSuccess Rate of De-HandingIncision QualitySelf-Adaptive Profiling Effect
Circumferential rotation methodFigure 560%66
Figure 6100%108
Figure 760%68
Axial plunge-cutting methodFigure 880%66
Figure 980%66
Figure 1060%66
Figure 1180%66
Figure 1280%88
Figure 13100%88
Figure 14100%108
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Guo, J.; Duan, J.; Yang, Z.; Karkee, M. De-Handing Technologies for Banana Postharvest Operations—Updates and Challenges. Agriculture 2022, 12, 1821. https://doi.org/10.3390/agriculture12111821

AMA Style

Guo J, Duan J, Yang Z, Karkee M. De-Handing Technologies for Banana Postharvest Operations—Updates and Challenges. Agriculture. 2022; 12(11):1821. https://doi.org/10.3390/agriculture12111821

Chicago/Turabian Style

Guo, Jie, Jieli Duan, Zhou Yang, and Manoj Karkee. 2022. "De-Handing Technologies for Banana Postharvest Operations—Updates and Challenges" Agriculture 12, no. 11: 1821. https://doi.org/10.3390/agriculture12111821

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop