







































 

 

 
145 

© 2025 Conscientia Beam. All Rights Reserved. 

Design, structural analysis, and ROS-based kinematic simulation of a robotic arm for capsicum 
harvesting in greenhouse environments   

 

 Ayan Paul1+ 

 Rajendra  
Machavaram2 

1,2Agricultural and Food Engineering Department, Indian Institute of 
Technology, Kharagpur, West Bengal, India. 
1Email: ayanpaul2210@kgpian.iitkgp.ac.in  
2Email: rajendra@agfe.iitkgp.ac.in 
  

(+ Corresponding author) 

 ABSTRACT 
 
Article History 
Received: 31 October 2025 
Revised: 1 December 2025 
Accepted: 3 December 2025 
Published: 5 December 2025 
 

Keywords 
6-DOF robotic arm 
Capsicum harvesting 
Forward kinematics 
ROS 2 simulation 
Structural analysis 
Workspace evaluation. 

 
This study presents an integrated mechanical design, structural validation, and kinematic 
simulation framework for a 6-DOF robotic manipulator tailored to precision capsicum 
harvesting in greenhouse environments. The robot was designed using a lightweight, 
modular architecture based on a hybrid-material approach: Polylactic Acid Plus (PLA+) 
structural links, Thermoplastic polyurethane (TPU) gripper fingers, and stainless-steel 
scissor cutters. The serial-link configuration was optimized for maneuvering within a 
canopy width and plant height range, with a total manipulator reach of approximately 
700 mm and a payload capacity of 200 g. Structural safety was verified via Finite Element 
Analysis (FEA) using SolidWorks Simulation, considering direction-specific harvesting 
loads. Maximum von Mises stress occurred in the shoulder joint (Link 2), while the 
forearm link (Link 5) showed the highest displacement, both remaining within allowable 
PLA+ limits. All links exhibited high Factors of Safety, with the minimum being 88.58, 
confirming mechanical integrity under worst-case loading. Kinematic modeling and 
motion planning were implemented using a URDF-based robotic model within ROS 2 
(Humble), RViz2, and MoveIt2. Forward kinematics showed a mean end-effector pose 
error of 2.3 mm across sampled target positions, validating model accuracy. The 
manipulator executed a full harvesting cycle in 6 seconds with sequential joint actuation, 
including base rotation, arm extension, wrist alignment, and gripping. Reachability and 
3D workspace analysis confirmed full coverage of the capsicum canopy (0.40 m radius) 
within a maximum reach of 0.70 m. The results demonstrate a biologically compatible, 
structurally safe, and kinematically feasible solution for greenhouse fruit harvesting. 
 

Contribution/Originality: This study contributes to the existing literature by integrating crop-specific 

morphology into robotic arm design for capsicum harvesting. It employs a new estimation methodology using link-

wise FEA and ROS 2-based kinematic validation. This study is among the few that have investigated PLA+ structural 

safety and full canopy reachability. 

 

1. INTRODUCTION 

Capsicum (Capsicum annuum L.), commonly known as sweet pepper or “Shimla Mirch,” is one of the most 

economically significant vegetable crops in India, with an annual production exceeding 500 thousand tonnes, 

primarily concentrated in states such as West Bengal, Karnataka, and Himachal Pradesh [1]. The increasing demand 

for high-quality produce and off-season cultivation has driven a significant shift toward protected cultivation systems. 

Greenhouse farming of capsicum has demonstrated yield improvements of 3–5 times over open-field conditions, while 

also enhancing fruit uniformity and quality. Nutritionally, capsicum is a valuable source of vitamin C (up to 130–190 

mg/100g), vitamin A, and antioxidants such as beta-carotene and flavonoids, making it highly desirable in both 

Current Research in Agricultural Sciences 
2025 Vol. 12, No. 2, pp. 145-164 
ISSN(e): 2312-6418 
ISSN(p): 2313-3716 
DOI: 10.18488/cras.v12i2.4572 
© 2025 Conscientia Beam. All Rights Reserved. 

 
 
 

 
 
 
 

 

 
 
 
 

https://orcid.org/0000-0002-4584-2119
https://orcid.org/0000-0002-1178-2015
mailto:ayanpaul2210@kgpian.iitkgp.ac.in
mailto:rajendra@agfe.iitkgp.ac.in
https://www.doi.org/10.18488/cras.v12i2.4572


Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
146 

© 2025 Conscientia Beam. All Rights Reserved. 

domestic and export markets [2]. However, the intensification of capsicum cultivation within controlled 

environments introduces several operational challenges, particularly in labor availability, harvesting efficiency, and 

postharvest handling. Given that capsicum fruits are non-climacteric and highly perishable, timely and precise 

harvesting is critical to preserving market quality and nutritional value. Manual harvesting, which remains the 

prevailing method, is not only labor-intensive but also susceptible to inconsistency and physical damage due to human 

error and fatigue. These limitations are further magnified in greenhouse settings, where the dense canopy, clustered 

fruit formation, and limited row spacing require high maneuverability and precision. In this context, automation 

through robotic harvesting systems presents a transformative solution. By leveraging advanced perception systems, 

motion planning algorithms, and specialized end-effectors, robotic platforms can effectively navigate the spatial 

constraints and mechanical complexities of greenhouse environments. Such systems offer the potential to reduce 

dependency on seasonal labor, standardize harvesting practices, minimize fruit loss, and enhance overall operational 

efficiency, thereby reinforcing the economic viability and scalability of protected agriculture. 

A typical robotic harvesting system for capsicum comprises several integrated subsystems working in 

coordination to enable autonomous operation. These include an RGB-D or multispectral vision module for real-time 

fruit detection and localization, a robotic manipulator responsible for accurate fruit detachment, a specialized end-

effector designed to perform non-destructive gripping and precise peduncle cutting, and a mobile platform capable of 

navigating the narrow and densely vegetated greenhouse rows [3]. Among these, the robotic manipulator serves as 

the central actuation unit, bridging the perception system with physical interaction [4]. Its ability to convert high-

level sensory inputs into precise motion control is critical for reaching and harvesting fruits situated at varying 

heights, orientations, and depths within the plant canopy, while also adhering to mechanical and safety constraints to 

prevent crop or structural damage. However, robotic harvesting in greenhouse environments presents several 

inherent challenges. Capsicum fruits frequently grow in clusters and are often partially or fully occluded by dense 

foliage, complicating visual detection and reachability. Moreover, the confined inter-row spacing, complex branching 

geometry, and limited working envelope further restrict manipulator motion. The stems of capsicum are also delicate 

and flexible, necessitating compliant and careful interaction to avoid bruising or plant injury. These constraints 

collectively emphasize the requirement for a robotic manipulator with a high number of degrees of freedom (DoF) 

and enhanced dexterity [5]. A multi-DoF arm, when integrated with a vision-guided control framework, enables 

adaptive path planning and orientation control to access occluded fruits, align the end-effector precisely with the 

peduncle, and execute harvesting actions with minimal collision risk. This motivates the design and development of 

a lightweight, perception-aware, and mechanically agile robotic arm tailored for greenhouse capsicum harvesting 

applications. 

Computer-Aided Design (CAD) serves as a fundamental tool in developing robotic arms for agricultural tasks by 

enabling precise joint configuration and geometric optimization for environments like greenhouses. Barth et al. [6] 

outlined structural and software considerations for agricultural robot subsystems using ROS middleware, 

highlighting its integration challenges and advantages. Barth et al. [6]. Rahul et al. [7] designed a 4-DOF parallel 

robot arm for automated seedling handling, achieving a 93.3% success rate with a 3.5 s cycle time, incorporating dual-

microcontroller firmware, stepper motor actuation, and 3D CAD-based mechanical modeling. Rahul et al. [7]. Van 

Herck et al. [8] demonstrated how CAD facilitates spatial planning for robotic systems in dense crop environments, 

emphasizing navigation and collision avoidance. Van Herck et al. [8]. Liming et al. [9] proposed a dual-arm robot 

using UR10 manipulators, employing standardized mechanical interfaces and ANSYS Workbench simulations to 

maintain strength and reduce weight to below 170 kg during live-wire operations. Liming et al. [9]. Huanca et al. 

[10] introduced “MARS-ROBOT,” an 8-DOF manipulator with a 1.0 m linear track, designed using SolidWorks and 

tested via MATLAB trajectory simulation and CoppeliaSim operability analysis for industrial painting tasks in Lima. 

Huanca et al. [10]. Paredes et al. [11] developed a pick-and-place and palletizing cell using SolidWorks and 

MATLAB-based kinematic modeling for SCARA T6 and UR10 robots, achieving task execution times of 1.18 s and 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
147 

© 2025 Conscientia Beam. All Rights Reserved. 

2.32 s, respectively. Paredes et al. [11]. Salazar and Portero [12] incorporated agricultural constraints such as torque 

and spatial limits into CAD-based 6-DOF robot arm models using Autodesk Inventor, enabling servo and gripper 

linkage design via iterative prototyping. Salazar and Portero [12]. Li and Li [13] introduced crop-robot co-design 

approaches, wherein CAD models aligned robot link design with plant canopy structures to improve reachability and 

perception. Li and Li [13]. Kuruvilla et al. [14] performed a comparative FEM analysis of a 6-DOF cylindrical arm 

for automated agriculture, evaluating stiffness across carbon fiber, aluminum, and steel variants under terrain slopes 

up to 60°, identifying cost-efficient, robust structures for precision farming [14]. 

Finite Element Analysis (FEA) enables evaluation of robotic structures under realistic stress and deformation 

scenarios, ensuring mechanical safety during repetitive agricultural operations. Zhang et al. [15] designed a 5-DOF 

hybrid serial-parallel manipulator using a novel 2T1R planar parallel mechanism, followed by FEA-based topology 

optimization to reduce structural weight while maintaining high stiffness and modularity, making it suitable for 

sorting and packaging tasks. Zhang et al. [15]. Amir et al. [16] implemented FEA for a strawberry-picking arm 

tailored to plant geometry, confirming mechanical safety and validating workspace simulations via MATLAB. Amir 

et al. [16]. Bauer et al. [17] applied structural analysis to multi-material, 3D-printed soft end-effectors for soil and 

crop interaction, validating their strength and durability despite low-cost fabrication. Bauer et al. [17]. Wang and 

Fan [18] introduced a bionics-inspired gripper with adjustable PVDF-based force sensing and validated its structure 

using FEA and Adams simulation, achieving a 23.2 ± 5% reduction in fruit damage during picking. Wang and Fan 

[18]. Li and Li [13] incorporated FEA-driven optimization into robotic arm design to enhance weight efficiency 

without compromising load capacity, optimizing link geometry under multi-axial loading and environmental stress. 

[14]. For capsicum harvesting robots, applying FEA under conditions such as grip force application, canopy collision, 

and long-term load cycles provides critical insights for reinforcement of joints, material selection (e.g., PLA+), and 

structural wall thickness calibration to prevent field failures. 

Kinematic modeling plays a pivotal role in robot motion planning, workspace validation, and collision-free 

trajectory generation, particularly in constrained agricultural settings. Wang et al. [19] developed a ROS-based 

simulation and control system for a dual-arm tomato harvesting robot, employing URDF for modeling, RViz and 

MoveIt! for motion planning, and integrating ROS-based vision and serial communication modules for precise 

localization Wang et al. [19]. Sepúlveda et al. [20] proposed a dual-arm ROS-compatible platform using 6-DOF 

Kinova MICO™ manipulators and RGB-Depth cameras, allowing environment-aware motion control and perception 

evaluation in unstructured farming scenarios Sepúlveda et al. [20]. Chen and Liu [21] implemented a ROS-based 

grasping robot system with real-time control via RViz, demonstrating high reliability and responsiveness in 

simulated and real part-handling tasks Chen and Liu [21]. Van Herck et al. [8] emphasized singularity avoidance 

and obstacle-aware motion planning in greenhouse robotics using forward and inverse kinematic models [8]. Modern 

approaches integrate kinematic modeling with real-time joint feedback from sensors like potentiometers or encoders 

to fine-tune pose estimation and improve simulation fidelity. ROS-based simulation environments like Gazebo or 

MoveIt! enable validation of joint trajectories, grasp-pose generation, and manipulability mapping before physical 

prototyping. These frameworks significantly reduce real-world trial needs, particularly when addressing occlusion, 

fruit clustering, and peduncle localization in capsicum harvesting. 

From the above literature, it is evident that although robotic systems for agricultural applications have seen 

significant advancements, there exists a specific gap in the mechanical and simulation-driven design of manipulators 

tailored for greenhouse capsicum harvesting. Current systems often generalize structural design without explicitly 

mapping the morphological traits of capsicum plants such as clustered fruit orientation, flexible peduncles, and narrow 

canopy architecture into design constraints, thereby limiting harvesting efficiency and safety. Additionally, while 

several robotic arms have been structurally analyzed, a detailed finite element analysis (FEA) of each link under fruit-

specific load conditions using lightweight 3D-printable materials like Polylactic Acid Plus (PLA+) remains 

underexplored. Furthermore, most simulation frameworks either lack ROS 2 integration or provide limited validation 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
148 

© 2025 Conscientia Beam. All Rights Reserved. 

of workspace coverage and motion planning strategies customized for capsicum picking. Hence, there is a pressing 

need for a comprehensive study that integrates mechanical CAD-based design, FEA-driven structural validation, and 

ROS 2-based kinematic simulation for a 6-DOF robotic arm optimized for capsicum harvesting in greenhouse 

environments. 

Based on this justification, the present study is structured around the following objectives. 

1. Mechanical Design: To design a lightweight, modular 6-DOF robotic arm for greenhouse capsicum harvesting, 

integrating plant-specific traits such as fruit height, canopy spread, stem flexibility, and occlusion into 

mechanical constraints. 

2. Structural Analysis: To perform FEA-based structural validation of the robotic arm under harvesting loads, 

ensuring safe operation by analyzing stress, deformation, and factor of safety for PLA+ components. 

3. Kinematic Simulation: To implement kinematic modeling and simulate motion planning using ROS 2, RViz2, 

and MoveIt2, validating reachability, joint trajectories, and workspace coverage for capsicum harvesting. 

This research adopts a structured methodology to address the design, validation, and simulation of a 6-DOF 

robotic arm tailored for capsicum harvesting in greenhouse environments. The paper begins with the Introduction, 

which outlines the growing need for automation in precision horticulture, identifies the morphological challenges 

posed by Capsicum annuum (e.g., clustered fruits, stem flexibility, occlusions), and highlights current research gaps in 

manipulator design and simulation-driven validation. Section 2: Materials and Methods presents a crop-informed 

engineering approach, starting from the greenhouse cultivation of the Arka Athulya capsicum variety to extract plant-

specific parameters for kinematic design. The section then details the CAD modeling of a modular robotic manipulator 

in SOLIDWORKS, followed by structural safety validation using Finite Element Analysis (FEA) under part-wise 

loading scenarios. It further introduces forward kinematic modeling using the Denavit–Hartenberg convention and 

elaborates on the ROS 2-based simulation framework using URDF file generation, MoveIt 2 for motion planning, 

and RViz 2 for trajectory visualization and workspace coverage analysis. Section 3: Results and Discussion presents 

the FEA output, including stress, displacement, and factor of safety results for PLA+ structures, followed by ROS 2 

simulation outcomes such as pose estimation accuracy, joint-angle time profiles during fruit approach and grasp, end-

effector reachability spheres, and coverage of capsicum harvesting zones. Gripper simplification for URDF export 

and actuation simulation is also addressed. Section 4: Conclusions summarizes the efficacy of the proposed 

manipulator design and ROS-based simulation pipeline, discusses its potential deployment readiness, and suggests 

directions for future enhancements. The paper concludes with a References section citing all supporting literature 

and simulation tools. 

 

2. MATERIALS AND METHODS 

2.1. Cultivation of Capsicum for Robotic Arm Design and Kinematics  

To establish a biologically grounded framework for robotic manipulator design, workspace evaluation, and 

inverse kinematic modeling, Capsicum annuum (cv. Arka Athulya) was cultivated under controlled conditions in a 

naturally ventilated greenhouse at the Agricultural and Food Engineering Department, Indian Institute of 

Technology Kharagpur (Figure 1). This experimental cultivation served as a critical step for acquiring detailed 

morphological, spatial, and mechanical data of the plant canopy essential for accurate robotic configuration. The Arka 

Athulya variety, characterized by uniform fruit morphology, sturdy peduncles, and a compact growth habit, was 

specifically chosen to support reliable robotic harvesting interactions. The sowing process began in October 2024, 

with seedlings initially raised in trays for a period of 30 days before transplanting onto structured raised beds 

arranged in a paired-row layout. A consistent inter-row spacing of 0.45 m and intra-row plant spacing of 0.40 m was 

maintained to establish a predictable canopy structure with defined occlusion zones and fruit distribution patterns. 

Raised beds measured 0.9 m in width and 0.15 m in height, with 1 m spacing between adjacent beds, ensuring sufficient 

clearance for robotic base mobility and arm articulation. The sandy loam soil type was selected to promote healthy 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
149 

© 2025 Conscientia Beam. All Rights Reserved. 

root development and maintain optimal plant physiology throughout the cropping cycle. This controlled and 

standardized cultivation setup (Table 1) enabled systematic capture of plant geometry and spatial variability, which 

directly informed the determination of joint range limits, reachable workspace volumes, and pose estimation 

parameters critical to the robotic arm's structural and kinematic design. 

 

 
Figure 1. Greenhouse cultivation of Capsicum annuum for robotic arm design. 

 

Table 1. Cultivation parameters for capsicum in greenhouse. 

Parameter Details 

Cultivar used Arka Athulya 
Fruit color Green 
Sowing method Seedlings raised in portrays 
Planting layout Paired row configuration 

Inter-row spacing 0.45 m 

Intra-row spacing 0.40 m 

Raised bed size (W × H) 0.9 m × 0.3 m 

Bed-to-bed spacing 1 m 
Soil type Sandy loam 
Cultivation period Winter (Nov 2024 – March 2025) 
Harvest window 75–85 days post-transplantation 
Total crop duration 5 – 6 months 

 

2.2. Design Considerations Based on Capsicum Plant Morphology 

Leveraging the structured greenhouse cultivation parameters outlined in Section 2.1, the robotic arm’s structural 

and kinematic configuration was systematically tailored based on the morphological characteristics and spatial 

organization of Capsicum annuum plants under real cultivation scenarios. These plant-specific attributes served as 

critical determinants in defining the manipulator’s geometric constraints, joint articulation strategy, and mechanical 

design limits. 

As presented in Table 2, essential geometric metrics including bed width (0.90 m), plant height range (0.60–

1.00 m), and canopy spread (0.30–0.60 m), directly influenced the robot’s reachable workspace and the required degree 

of freedom (DOF) configuration. The narrow inter-plant spacing (0.40 m) and clustered canopy structure necessitated 

a compact arm footprint with enhanced lateral flexibility, optimized for integration onto mobile platforms navigating 

confined greenhouse aisles. The fruit-bearing zone, typically located 0.30–0.80 m above the ground, dictated vertical 

reachability and end-effector trajectory constraints, especially when addressing occluded or partially visible fruit 

clusters containing 2–5 fruits per node. Payload considerations, derived from the average fruit mass (120–180 g), 

informed actuator sizing and gripper compliance thresholds to ensure efficient yet non-damaging harvesting 

operations. 

Further, the mechanical interaction dynamics during harvesting were defined through an evaluation of stem and 

peduncle biomechanics, detailed in Table 3. The stem exhibits moderate stiffness and elastic recovery, which 

necessitates a compliant and adaptive motion profile across joints to prevent structural damage to the plant. Given 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
150 

© 2025 Conscientia Beam. All Rights Reserved. 

the constraint that maximum permissible contact forces must remain below 10 N, fine-tuned force control and real-

time collision avoidance become imperative for safe operation. These biomechanical insights guided critical design 

decisions such as lightweight material selection, joint torque specifications, and smooth trajectory generation. 

Collectively, these parameters informed the development of a biologically compatible and functionally robust robotic 

arm capable of autonomous capsicum harvesting in spatially constrained greenhouse environments. 

 

Table 2. Geometric and structural parameters of capsicum plants relevant to robotic manipulator design. 

Parameter Value Design Implication 

Greenhouse bed width 90 cm 
Constrains the base frame and lateral mobility of the robot; impacts 
mounting feasibility. 

Plant height 60–100 cm 
Sets the minimum vertical reach requirement to access fruit across 
canopy layers. 

Canopy spread (Width 
& depth) 

30–60 cm 
Defines operational workspace boundaries for the end-effector; 
critical for collision-free movement. 

Intra-row plant 
spacing 

40 cm 
Demands compact link dimensions and a reduced turning radius for 
maneuverability in narrow gaps. 

Branch count and 
growth pattern 

6–10 branches, 
radial orientation 

Influences joint configuration for collision avoidance and trajectory 
optimization. 

Fruit height from 
ground 

30–80 cm 
Guides end-effector positioning and approach angles for efficient 
harvesting. 

Average fruit mass 120–180 g 
Used for payload estimation, actuator torque selection, and gripper 
force calibration. 

Occlusion level of 
peduncle/Fruit 

High in dense 
canopy 

Necessitates vision-based pose estimation and multi-DOF arm 
articulation. 

Fruit cluster density 
2–5 fruits per 
node 

Demands precision manipulation and high dexterity to avoid 
damage during selective picking. 

 

Table 3. Mechanical interaction characteristics of capsicum plants influencing manipulator design. 

Parameter Value Design Implication 

Stem and branch 
flexibility 

Medium to high 
Requires compliant motion strategies and joint-level adaptability 
to avoid plant damage. 

Max. permissible 
contact force 

< 10 N 
Dictates end-effector force thresholds for safe fruit engagement 
and detachment. 

Stiffness and recovery 
response 

Moderate with fast 
rebound 

Informs damping design and control logic for safe impact 
handling and responsive motion control. 

 

2.3. Structural Design of the Robotic Manipulator  

Taking into account the spatial limitations and biological characteristics of capsicum plants outlined in Section 

2.2, a six-degree-of-freedom (6-DOF) articulated robotic manipulator was structurally designed using 

SOLIDWORKS Premium 2022 (Dassault Systèmes, Vélizy-Villacoublay, France) [22]. The design prioritized 

lightweight construction, modularity, and the ability to effectively navigate within greenhouse canopies. The robot 

adopts a serial-link configuration consisting of seven primary links, including the wrist and end-effector, with joint 

trajectories optimized for operation within a confined canopy spread of 30–60 cm and plant height envelope of 60–

100 cm (Figure 2). For prototyping, the structural links were fabricated via Fused Deposition Modeling (FDM) using 

Polylactic Acid Plus (PLA+), selected for its balance of stiffness and ease of manufacturing. The end-effector employed 

a hybrid-material strategy: gripper elements were constructed from flexible Thermoplastic polyurethane (TPU) 

(Shore hardness 85A) to ensure compliant fruit engagement, while the integrated scissor-based cutting mechanism 

was fabricated from stainless steel to enable reliable peduncle severing. This multi-material approach enabled a low-

mass manipulator with a rated payload capacity of 200 g, adequate for handling capsicum fruits typically weighing 

between 120–180 g, along with the attached end-effector assembly. The dimensional specifications and material 

allocation for each robotic link are summarized in Table 4, based on the final prototype measurements. 

 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
151 

© 2025 Conscientia Beam. All Rights Reserved. 

Table 4. Link-wise dimensions, materials, and weights of the capsicum harvesting robotic arm. 

Link No. Function Material Length (mm) Weight (kg) 

Link 1 Base mounting PLA+ 70 0.45 
Link1_wheel Support above base PLA+ 20 0.15 
Link 2 Shoulder joint PLA+ 200 0.50 
Link 3_a Upper arm (Segment A) PLA+ 140 0.40 
Link 3_b Upper arm (Segment B) PLA+ 190 0.35 
Link 4 Elbow joint PLA+ 175 0.30 
Link 5 Forearm PLA+ 130 0.10 
Link 6 Wrist pitch/Roll PLA+ 75 0.05 
Link 7 End-effector + cutter PLA+/TPU/SS 130 0.45 

 

 
Figure 2. CAD model of the robotic arm with end-effector for capsicum harvesting. 

 

2.4. Structural Analysis of Designed Robotic Manipulator 

Building upon the structural configuration described in Section 2.3, the mechanical robustness of the designed 6-

DOF articulated robotic arm was evaluated through Finite Element Analysis (FEA) using SolidWorks Simulation 

(Premium 2022, Dassault Systèmes, France) [22]. FEA enables the prediction of stress distribution (Equation 1), 

deformation (Equation 2), and structural performance under operational loading conditions, which is critical for 

validating the safety (Equation 3) and reliability of the robotic system prior to deployment. SolidWorks Static 

Simulation module was employed to assess the structural integrity of each individual link rather than the entire 

assembly to allow for isolated evaluation under relevant cumulative loading, simplify constraint definition, and reduce 

meshing complexity. The simulation setup specifications are summarized in Table 5, which outlines solver type, mesh 

strategy, and analysis metrics such as Von Mises stress and total displacement. 

σvm = √
(σ1− σ2)2+ (σ2− σ3)2+ (σ3− σ1)2

2
      (1) 

 𝛿 =  
𝐹⋅𝐿

𝐴⋅𝐸 
    (2) 

𝐹𝑜𝑆 =  
𝜎𝑚𝑎𝑥

𝜎𝑦𝑖𝑒𝑙𝑑 
     (3) 

Where: 

• σvm = Von Mises Stress. 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
152 

© 2025 Conscientia Beam. All Rights Reserved. 

• δ = Total Linear Deformation. 

• FoS = Safety Factor. 

• 𝜎1,  𝜎2,  𝜎3  = Principal stresses. 

• δ = Linear deformation (m). 

• F = Applied axial force (N). 

• A= Cross-sectional area (m²). 

• L = Length of the link (m). 

• E = Young’s modulus of the material (Pa). 

• 𝜎𝑦𝑖𝑒𝑙𝑑 = Yield strength of material (Pa). 

• 𝜎𝑚𝑎𝑥  = Maximum simulated Von Mises stress (Pa). 

The FEA process consisted of four key steps for each link: (i) assignment of appropriate material properties based 

on the component function and manufacturing feasibility, (ii) application of fixture constraints to simulate physical 

mounting or joint anchoring conditions, (iii) application of external loads computed from cumulative mass above each 

link, and (iv) meshing using the Standard Mesh algorithm with curvature-based refinement for improved accuracy. 

To ensure realistic simulation fidelity, material definitions were applied based on actual component fabrication, 

as shown in Table 6. Rigid links (base to wrist) were assigned PLA+ due to its lightweight nature and acceptable 

strength-to-weight ratio for 3D-printed components, while the compliant fingers of the end-effector were modeled 

with TPU (Shore 85A) for flexibility. The cutting interface utilized stainless steel to withstand shearing forces during 

peduncle removal.  

Material parameters, including elastic modulus, yield strength, Poisson’s ratio, and density were incorporated 

into the simulation to model accurate deformation and failure thresholds. The mechanical loading was determined 

through a bottom-up approach, accounting for cumulative weight and force transmission through the robotic chain 

during harvesting operations.  

As detailed in Table 7, individual components such as Link 2 (shoulder), Link 3 (arm segments), and Link 8 (end-

effector) were subjected to their corresponding axial and lateral loads derived from gravitational force calculations. 

Load directionality was aligned with expected working conditions: vertically downward for gravity loads and 

inward/outward vectors based on actuation axis and fruit access strategy. Boundary conditions (fixtures) were 

assigned at realistic mounting faces (e.g., fixed base flange, joint-to-joint interfaces), and standard meshing was 

employed for reliable convergence. 

This structural analysis pipeline validated that all robotic arm components fall within acceptable stress limits 

under expected harvesting loads, and guided iterative reinforcement of critical links such as the forearm and wrist. 

These insights directly informed material choices and design tolerances, ensuring that the robotic arm not only meets 

spatial and kinematic constraints from Section 2.2 but also withstands the mechanical demands of greenhouse 

harvesting. 

 

Table 5. Software setup for static simulation in SOLIDWORKS. 

Parameter Specification 

Cad software SOLIDWORKS Premium 2022 

Simulation type Static Structural 

Gravity enabled Yes (9.81 m/s²) 

Mesh type Standard Mesh (Curvature-based) 

Solver FFEPlus (iterative) 

Result metrics Von Mises stress, displacement, FOS 

 

 

 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
153 

© 2025 Conscientia Beam. All Rights Reserved. 

Table 6. Material properties used for FEA simulation. 

Property PLA+ TPU (Shore 85A) Stainless Steel (304) 

Young’s Modulus (MPa) 3450 26 200,000 
Poisson’s Ratio 0.36 0.48 0.30 
Density (kg/m³) 1240 1200 8000 
Yield Strength (MPa) 50 7.89 250 
Ultimate Tensile Strength (MPa) 65 8.5 505 
Thermal Expansion (1/K) 68e-6 160e-6 17.2e-6 
Damping Ratio (Structural) Low Moderate Low 
Behavior Rigid-Plastic Hyperelastic-like Rigid-Elastic 

 

Table 7. Link-wise simulation setup for FEA in SOLIDWORKS. 

Link 
no. 

Component name Material 
Applied 
force (N) 

Load direction Fixture condition 

1 Link 1 (Base) PLA+ 26.98 
Vertical 
downward  

Fixed at bottom face 
(Mount to platform) 

2 
Link 1_w (Wheel 
after Base) 

PLA+ 22.56 
Vertical 
downward  

Fixed concentrically to 
base flange. 

3 Link 2 (Shoulder) PLA+ 21.09 
Downward & 
inward  

Fixed at joint with Link 1 

4 
Link 3_a (Lower 
Arm) 

PLA+ 16.19 
Downward & 
inward  

Fixed to Link 2 interface 

5 
Link 3_b (Upper 
Arm) 

PLA+ 12.26 
Downward & 
inward  

Fixed to Link 3_a 
connection 

6 Link 4 (Elbow) PLA+ 8.83 
Downward & 
inward  

Fixed to Link 3_b joint 

7 Link 5 (Forearm) PLA+ 5.89 
Downward & 
inward  

Fixed to Link 4 
connection 

8 Link 6 (Wrist) PLA+ 4.91 
Downward & 
inward  

Fixed to Link 5 
connection 

9 
Link 8 (End-
effector) 

PLA+/TPU/SS 4.41 
Downward & 
outward  

Fixed at wrist interface 
(Link 6) 

 

2.5. Kinematic Modeling and ROS 2-Based Simulation Framework 

To validate the reachability, workspace constraints, and joint-level articulation of the designed capsicum 

harvesting manipulator, forward kinematic modeling and ROS 2-based simulation were implemented. Forward 

kinematics provides the pose of the end-effector (position and orientation) given the joint parameters and is 

fundamental in evaluating the manipulator's capability to reach spatial targets under joint and structural constraints. 

This was essential for confirming whether the robotic arm could effectively navigate the wide canopy and tall plants 

described earlier in Section 2.3. The Denavit–Hartenberg (DH) parameterization method was used to model the 6-

DOF serial-link configuration, where each transformation between two consecutive joints was defined by four 

parameters: link length, link twist, link offset, and joint angle. The complete DH parameter set used in this study is 

summarized in Table 8.  

The simulation architecture was built on ROS 2 Humble [23] using RViz2 and MoveIt2 for motion planning. 

The robotic arm’s mechanical structure, defined through the SolidWorks CAD model, was converted into a Unified 

Robot Description Format (URDF) (Figure 3) to enable integration with ROS tools. Each link was exported as an 

STL mesh and referenced in the URDF using and tags. The inertial data were calculated from link-wise physical 

properties obtained through Section 2.4’s FEA analysis. All joints were defined as revolute with appropriate axis 

orientation and motion limits, derived from task-specific velocity requirements and load-based constraints. The joint-

level actuation specifications critical for motion planning are provided in Table 9, including the functional purpose, 

practical RPM range, and rationale for each joint’s design. 

 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
154 

© 2025 Conscientia Beam. All Rights Reserved. 

 
Figure 3. Robotic arm URDF structure. 

 

To simulate robot behavior and test kinematic feasibility, a new ROS 2 package named capsicum_arm_description 

was created within the workspace ws_gazebo, containing the URDF, Xacro, and mesh resources. Launch files were 

developed using robot_state_publisher, joint_state_publisher_gui, and RViz2 configurations to visualize and 

manually actuate the robotic joints. The forward kinematics and collision-free path validation were achieved through 

MoveIt2, which was initialized using the MoveIt Setup Assistant. This setup allowed the definition of planning 

groups, end-effectors, kinematic solvers, and motion constraints specific to the greenhouse harvesting context. The 

final MoveIt2 planning interface supported inverse kinematics computations, joint-space planning, and Cartesian path 

generation. Importantly, joint constraints and workspace boundaries defined from real plant geometry were encoded 

in the SRDF collision matrix and planning configuration, ensuring biologically compatible motion.  

 

Table 8. Denavit–Hartenberg Parameters for the 6-DOF robotic manipulator. 

Joint 𝒂𝒊  (mm) 𝛂𝒊 (deg) 𝒅𝒊 (mm) 𝛉𝒊 (variable) 

J1 0 90 65 θ1 

J2 200 0 0 θ2 

J3 140 0 0 θ3 

J4 190 90 0 θ4 

J5 175 90 0 θ5 

J6 130 0 0 θ6 

 

Table 9. Joint specifications and RPM constraints for simulation and motion planning. 

Joint (Link) Joint Type Function RPM Reason 

Base (J1) Revolute 
Yaw rotation of the full 
arm 

10–20 Enables stable sweeping motion of the full arm. 

Shoulder (J2) Revolute Main lifting arm 10–30 Requires more torque and slower, controlled lift. 
Elbow (J3) Revolute Bending forearm 15–35 Allows faster arm folding for reach. 
Wrist 1 (J4) Revolute Pitching of wrist 20–50 Lightweight, allows rapid orientation. 
Wrist 2 (J5) Revolute Rolling of wrist 20–50 Aids in dexterous manipulation 

Gripper (J6) Revolute 
Scissor/cutting 
mechanism 

30–60 Requires fast, responsive actuation. 

 

Table 10. ROS 2 and simulation tool specifications. 

Tool Version/Specification Function 

ROS 2 Humble Hawksbill (Ubuntu 22.04) Core middleware for robotic system integration 
RViz2 Default with ROS 2 Humble 3D visualization of robot state and sensors 
MoveIt2 Binary install from ROS 2 repo Motion planning, FK/IK, collision detection 
URDF/Xacro XML and macro format Robot description and modular link modeling 
STL Files Exported from SolidWorks Link geometry used for visual/collision modeling 
Joint State GUI ROS GUI tool Manual testing of joint limits and articulation 

 

The complete ROS 2–MoveIt2–RViz2 pipeline enabled real-time simulation of peduncle-approach trajectories, 

reachability analysis, and motion validation without the need for physical trials. The specifications for ROS 2 packages 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
155 

© 2025 Conscientia Beam. All Rights Reserved. 

and tools used in this simulation setup are consolidated in Table 10, establishing the simulation stack as a reproducible 

and scalable framework for field-ready agricultural robotic systems. 

This integrated simulation architecture ensures that the designed manipulator can be rigorously evaluated in a 

virtual environment before hardware deployment, with joint-level motion planning, spatial reach validation, and real-

time visualization aligned with the biological and mechanical constraints identified in Sections 2.2 to 2.4. 

 

3. RESULTS AND DISCUSSION 

3.1. FEA Analysis Results of Designed Robotic Manipulator 

Following the structural modeling and mechatronic integration methodology described in Section 2.3, Finite 

Element Analysis (FEA) was performed on each individual link of the robotic manipulator to evaluate mechanical 

safety under operational loads. The simulations were conducted using SolidWorks Simulation (Premium 2022, 

Dassault Systèmes, France), leveraging the static analysis environment with standard meshing settings and isotropic 

linear material properties corresponding to PLA+ for primary structural links, and TPU and Stainless Steel for the 

gripper segment. The analysis was carried out independently for each link, as opposed to full-assembly simulation, to 

improve computational efficiency and isolate stress distribution profiles under localized loading, based on the vertical 

and directional load values previously calculated in Table 8. 

The FEA results are consolidated in Table 11, which reports the von Mises stress, total displacement, and Factor 

of Safety (FoS) for all nine primary components of the manipulator. The highest stress concentration was observed 

in Link 2, which carries the largest cumulative load (21.09 N), registering a maximum stress of 5.644 × 10⁵ Pa. This 

result aligns with the role of Link 2 as the shoulder joint, a critical load-bearing element that not only supports upper 

links but also handles significant bending torque during vertical movements. In contrast, the lowest stress of 8.529 

× 10³ Pa was found in Link 1_wheel, a stationary support component that contributes minimally to motion or payload 

distribution. 

Regarding displacement, the highest value was recorded in Link 5 at 2.53 × 10⁻² mm, likely due to its position at 

the distal end of the manipulator where accumulated flexural deflection from upper links is maximum. This 

displacement remains well within allowable deformation limits for PLA+, suggesting no risk of permanent structural 

deviation under operating conditions. The lowest displacement of 3.460 × 10⁻⁵ mm occurred in Link 1_wheel, which 

is fixed to the base with minimal external loading and experiences negligible stress propagation. 

The Factor of Safety (FoS) values across all parts significantly exceeded critical thresholds, with the minimum 

FoS still being 88.58, again in Link 2, further confirming its criticality under load. Despite being the most stressed 

part, this FoS confirms a substantial safety margin when compared to the yield strength of PLA+ (reported as 50 

MPa). These high FoS values are attributable to conservative material usage and the lightweight, modular design 

intended for low-payload capsicum harvesting operations. Emphasis was placed on PLA+ components due to their 

relatively lower stiffness and strength compared to metal counterparts. This allowed validation of the robot’s 

structural feasibility even under worst-case scenarios, highlighting that even in these critical parts, material failure is 

highly unlikely. Stress and displacement contour plots for each component are illustrated in Table 11, with high-

stress zones observed near joint interfaces, particularly in torque-sensitive links such as Link 2 and Link 3_a, 

emphasizing the importance of reinforcement in design. 

Hence, the FEA results confirm that the robotic manipulator, as designed and fabricated using PLA+ material, 

meets all structural safety requirements under expected loading conditions for capsicum harvesting. These insights 

guide future enhancements in actuator selection, material optimization, and topology refinement for advanced 

iterations of the system. 

 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
156 

© 2025 Conscientia Beam. All Rights Reserved. 

Table 11. FEA results for each manipulator link showing von Mises stress, displacement, and factor of safety. 

Link Stress Plots Displacement Plots FoS Plots 

Link 
1 

   

Link 
1_w 

   

Link 
2 

  
 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
157 

© 2025 Conscientia Beam. All Rights Reserved. 

Link 
3_a 

  
 

Link 
3_b 

 
 

 

Link 
4 

  
 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
158 

© 2025 Conscientia Beam. All Rights Reserved. 

Link 
5 

   

Link 
6 

   

Link 
7 

  
 

 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
159 

© 2025 Conscientia Beam. All Rights Reserved. 

3.2. ROS 2-Based Simulation Results 

To evaluate the performance, reachability, and practical motion feasibility of the developed 6-DOF harvesting 

manipulator, a detailed simulation study was conducted based on the ROS 2 framework, as described in Section 2.5. 

The robotic model was implemented using URDF derived from the CAD geometry and simulated within MoveIt2 

and RViz2 environments. For improved compatibility with simulation tools and to streamline the exportation process, 

the original end-effector with an integrated scissor mechanism was simplified into a parallel-jaw gripper. This 

abstraction preserved the essential grasping function while minimizing complexity in mesh handling and frame 

association during motion planning. Figure 4 shows the planning scene as visualized in RViz2, where the manipulator 

model is rendered with proper joint hierarchy, link geometries, and interactive planning markers. The coordinate 

frames for each joint are correctly aligned, confirming successful construction of the kinematic chain. The MoveIt2 

interface was used to configure motion groups, validate forward kinematics, and generate collision-free paths between 

target poses. The gripper, represented with a simplified closing mechanism, was able to simulate basic approach and 

pick actions needed for harvesting tasks. 

The actuator-level behavior of the robot was further analyzed through the joint angle evolution over time, 

reflecting a full motion cycle from base rotation to fruit grasping. Figure 5 illustrates the angular displacement of 

each revolute joint (θ₁ to θ₆) over a 6-second period. The trajectory begins with θ₁ executing a base rotation of 

approximately 45°, positioning the arm toward the target location. This is followed by a shoulder pitch (θ₂) and elbow 

extension (θ₃), which lift and extend the arm into the workspace. Notably, the joints are activated in a sequential and 

phase-wise manner, indicating task-specific decoupling that reduces dynamic interaction and control complexity. The 

wrist joints (θ₄ and θ₅) contribute to orienting the end-effector appropriately, while the final joint (θ₆), which controls 

the gripper, activates rapidly at around 5 seconds coinciding with the moment of fruit capture. The joint profiles 

verify that the manipulator operates within its defined kinematic and dynamic constraints while executing a realistic 

harvesting sequence. 

 

 
Figure 4. ROS 2 RViz visualization of the robotic arm model with simplified end-effector in MoveIt2. 

 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
160 

© 2025 Conscientia Beam. All Rights Reserved. 

 
Figure 5. Joint angle profiles for the full 6-DOF manipulator during a single harvesting motion cycle. 

 

To evaluate the positional fidelity of the forward kinematics model, 20 spatial poses were sampled across the 

robot’s reachable workspace. The position error between the computed end-effector pose and the expected target was 

recorded, as shown in Figure 6. The error varied between 1.8 mm and 2.8 mm, with a mean value around 2.3 mm. 

These small deviations demonstrate that the DH-parameter-based transformation model used in the simulation 

closely approximates the true spatial configuration, ensuring that the end-effector can be reliably positioned within 

biologically tolerable limits. The observed variation is attributable to minor mesh misalignments and cumulative 

transformation inaccuracies but remains within acceptable thresholds for agricultural applications, where fruit 

peduncles typically allow for a few millimeters of positional flexibility. 

 

 
Figure 6. End-effector position error from forward kinematics across 20 target pose samples. 

 

The planar reachability of the manipulator was visualized using a normalized polar plot as depicted in Figure 7. 

The reachability envelope represents the maximum radial extent the end-effector can achieve at various angles. The 

shape of the envelope is slightly asymmetric, reflecting the influence of joint angle limits and differing link lengths. 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
161 

© 2025 Conscientia Beam. All Rights Reserved. 

Despite these variations, the manipulator exhibits near-complete circular coverage, demonstrating its potential to 

access targets in all lateral directions. Such comprehensive reach is vital for operations in irregular and multi-layered 

plant canopies where fruits may appear at varying azimuthal positions. 

 

 
Figure 7. End-effector polar reachability envelope visualized in normalized radial units. 

 

In three-dimensional space, the extent of the manipulator’s reachable volume was compared with the expected 

fruit-bearing canopy coverage typical of capsicum plants. Figure 8 shows this volumetric overlay, with the outer blue 

region representing the full manipulator reach (0.70 m) and the inner green region denoting the plant canopy 

coverage (0.40 m). The spatial overlap confirms that the robot is not only capable of reaching all desired fruit locations 

but also has additional margin for obstacle avoidance and approach optimization. This analysis establishes that the 

manipulator’s workspace has been well-matched to the biological domain it is designed to serve. 

Collectively, these results confirm that the proposed manipulator design, its simplified end-effector configuration, 

and the ROS 2-based simulation framework form a reliable and biologically compatible platform for greenhouse fruit 

harvesting. The integration of mechanical, kinematic, and computational considerations within this simulation 

pipeline allows for effective pre-deployment testing, reducing the risk and cost associated with field-level trials. 

 

 
Figure 8. Workspace comparison showing total manipulator reachability and capsicum canopy coverage. 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
162 

© 2025 Conscientia Beam. All Rights Reserved. 

4. CONCLUSIONS 

This research presents an integrated framework for the design, structural validation, and motion planning 

simulation of a 6-DOF robotic arm tailored for capsicum harvesting in structured greenhouse environments. Firstly, 

a modular and lightweight manipulator was designed based on the spatial and morphological characteristics of 

capsicum plants, such as canopy spread, fruit positioning, stem flexibility, and occlusion patterns, ensuring crop-

structure compatibility in mechanical design. Then, finite element analysis was conducted on each link to assess stress 

distribution, deformation, and factor of safety under representative harvesting loads, confirming the structural 

integrity of the PLA+ and multi-material components. Finally, a forward kinematic model was developed using the 

Denavit–Hartenberg convention and integrated into a ROS 2-based simulation environment. The robot was modeled 

in URDF and visualized in RViz2 and MoveIt2 to evaluate motion feasibility, workspace coverage, and joint 

trajectories. The generated results verified that the system can safely and effectively operate within greenhouse 

constraints, demonstrating its suitability for precision horticultural automation. 

The following key conclusions summarize the technical achievements of the study. 

1. A 6-DOF articulated robotic arm was mechanically designed using a hybrid-material approach involving 

PLA+, TPU (Shore hardness 85A), and stainless steel to meet the spatial and structural requirements of 

greenhouse-grown capsicum plants. The serial-link configuration was optimized for maneuvering within a 

canopy width of 30–60 cm and plant height of 60–100 cm, while accommodating a payload capacity of 200 g. 

The final prototype consisted of 9 major links with individual link lengths ranging from 20  mm (wheel) to 

200 mm (shoulder), resulting in an overall reach of approximately 700 mm. The lightweight design was 

achieved through FDM-based 3D printing using PLA+ with link weights between 0.05–0.50 kg. The end-

effector combined flexible TPU gripper fingers with a stainless-steel scissor for peduncle cutting. 

2. Finite Element Analysis (FEA) was conducted independently on each robotic link under direction-specific 

operational loads using SolidWorks Simulation (Premium 2022, Dassault Systèmes, France). The highest von 

Mises stress was observed in Link 2 at 5.644 × 10⁵ Pa, corresponding to its role as the main load-bearing 

shoulder segment. The lowest stress occurred in Link 1_wheel at 8.529 × 10³ Pa. Maximum displacement was 

noted in Link 5 at 2.53 × 10⁻² mm, and minimum in Link 1_wheel at 3.460 × 10⁻⁵ mm. The minimum factor of 

safety (FoS) was 88.58, again in Link 2, well above the critical threshold relative to PLA+ yield strength 

(50 MPa), confirming mechanical robustness under worst-case harvesting loads. 

3. The ROS 2-based simulation environment using RViz2 and MoveIt2 validated the robot’s reachability, pose 

tracking, and motion feasibility. The URDF model accurately replicated forward kinematics with a mean end-

effector position error of 2.3 mm across 20 pose samples. Joint angle profiles demonstrated sequential motion 

from θ₁ to θ₆ over a 6-second harvesting cycle, with θ₁ rotating 45° and θ₆ activating at 5 s for gripping. The 

polar reachability envelope confirmed a circular planar workspace with a maximum radial reach of 0.70 m. A 

3D workspace comparison showed complete overlap with the capsicum canopy volume (0.40 m radius), 

ensuring full accessibility of fruit locations with a margin for path optimization. 

The study thus successfully delivers a fully integrated design, structural validation, and kinematic simulation 

framework for a 6-DOF robotic arm tailored to high-precision, non-destructive capsicum harvesting within 

greenhouse environments. The proposed system demonstrates strong mechanical reliability, optimal reachability for 

the capsicum canopy space, and high kinematic accuracy using ROS 2-based simulations. While the robotic 

manipulator achieved biologically relevant pose accuracy and complete canopy coverage in simulation, current 

validation was limited to static targets and controlled conditions within a virtual environment. Future work will 

target real-time deployment under variable fruit occlusion and foliage complexity. Integration with mobile bases, 

low-power actuators, and trajectory optimization strategies will further enhance the platform’s adaptability for large-

scale horticultural operations. Moreover, incorporating adaptive motion planning under multi-object occlusion is 



Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
163 

© 2025 Conscientia Beam. All Rights Reserved. 

expected to improve system generalization, decision-making under uncertainty, and the real-world applicability of 

the proposed harvesting framework. 

 

Funding: This research was supported by the Indian Institute of Technology, Kharagpur, West Bengal (Grant 
number: PMRF 2402325). 
Institutional Review Board Statement: Not applicable. 
Transparency: The authors state that the manuscript is honest, truthful, and transparent, that no key aspects 
of the investigation have been omitted, and that any differences from the study as planned have been clarified. 
This study followed all writing ethics. 
Competing Interests: The authors declare that they have no competing interests. 
Authors’ Contributions: Both authors contributed equally to the conception and design of the study. Both 
authors have read and agreed to the published version of the manuscript. 
Disclosure of AI Use: During the preparation of this work, I utilized ChatGPT (OpenAI, GPT-4o model, 
accessed via https://chat.openai.com) solely for enhancing the language clarity of certain sections. The tool 
was not used for content generation, data analysis, or drawing scientific conclusions. After using this service, 
I thoroughly reviewed and edited the content as necessary, and I take full responsibility for the final content 
of the published article. 

 

REFERENCES 

[1] A. Paul, R. Machavaram, Ambuj, D. Kumar, and H. Nagar, "Smart solutions for capsicum Harvesting: Unleashing the 

power of YOLO for detection, segmentation, growth stage classification, counting, and real-time mobile identification," 

Computers and Electronics in Agriculture, vol. 219, p. 108832, 2024.  https://doi.org/10.1016/j.compag.2024.108832 

[2] A. Paul and R. Machavaram, "Greenhouse capsicum detection in thermal imaging: A comparative analysis of a single-

shot and a novel zero-shot detector," Next Research, vol. 1, no. 2, p. 100076, 2024.  

https://doi.org/10.1016/j.nexres.2024.100076 

[3] A. Kaleem, S. Hussain, M. Aqib, M. J. M. Cheema, S. R. Saleem, and U. Farooq, "Development challenges of fruit-

harvesting robotic arms: A critical review," AgriEngineering, vol. 5, no. 4, pp. 2216-2237, 2023.  

https://doi.org/10.3390/agriengineering5040136 

[4] M. Xie, Fundamentals of robotics: linking perception to action. Singapore: World Scientific Publishing Company, 2003.  

[5] Y. Fan, X. Lv, J. Lin, J. Ma, G. Zhang, and L. Zhang, "Autonomous operation method of multi-DOF robotic arm based 

on binocular vision," Applied Sciences, vol. 9, no. 24, p. 5294, 2019.  https://doi.org/10.3390/app9245294 

[6] R. Barth et al., "Using ROS for agricultural robotics-design considerations and experiences," in Proceedings of the Second 

International Conference on Robotics and Associated High-Technologies and Equipment for Agriculture and Forestry, 2014, pp. 

509-518.  

[7] K. Rahul, H. Raheman, and V. Paradkar, "Design of a 4 DOF parallel robot arm and the firmware implementation on 

embedded system to transplant pot seedlings," Artificial Intelligence in Agriculture, vol. 4, pp. 172-183, 2020.  

https://doi.org/10.1016/j.aiia.2020.09.003 

[8] L. Van Herck, P. Kurtser, L. Wittemans, and Y. Edan, "Crop design for improved robotic harvesting: A case study of 

sweet pepper harvesting," Biosystems Engineering, vol. 192, pp. 294-308, 2020. 

https://doi.org/10.1016/j.biosystemseng.2020.01.021 

[9] Z. Liming, H. Yulong, X. Shanjun, Z. Tong, G. Junlong, and W. Mingrui, "Mechanical design and finite element analysis 

of live working robot for 10kV distribution power systems," Procedia Computer Science, vol. 183, pp. 331-336, 2021.  

https://doi.org/10.1016/j.procs.2021.02.067 

[10] J. Huanca, J. Zamora, J. Cornejo, and R. Palomares, "Mechatronic design and kinematic analysis of 8 DOF serial robot 

manipulator to perform electrostatic spray painting process on electrical panels," presented at the 2022 IEEE 

Engineering International Research Conference (EIRCON), IEEE, 2022.  

[11] C. Paredes, R. Palomares, J. Alva, and J. Cornejo, "Mechatronics design and robotic simulation of serial manipulators to 

perform automation tasks in the avocado industry," International Journal of Advanced Computer Science and Applications, 

vol. 14, no. 8, pp. 509–517, 2023. https://doi.org/10.14569/IJACSA.2023.0140856 

https://doi.org/10.1016/j.compag.2024.108832
https://doi.org/10.1016/j.nexres.2024.100076
https://doi.org/10.3390/agriengineering5040136
https://doi.org/10.3390/app9245294
https://doi.org/10.1016/j.aiia.2020.09.003
https://doi.org/10.1016/j.biosystemseng.2020.01.021
https://doi.org/10.1016/j.procs.2021.02.067
https://doi.org/10.14569/IJACSA.2023.0140856


Current Research in Agricultural Sciences, 2025, 12(2): 145-164 

 

 
164 

© 2025 Conscientia Beam. All Rights Reserved. 

[12] M. Salazar and P. Portero, "CAD design and modeling of a robotic Arm for automated harvesting," in Proceedings of the 

International Conference on Robotics and Automation, IEEE, 2024, pp. 123–130.  

[13] H. Li and Y. Li, "Finite element analysis and structural optimization design of multifunctional robotic arm for garbage 

truck," Frontiers in Mechanical Engineering, vol. 11, p. 1543967, 2025.  https://doi.org/10.3389/fmech.2025.1543967 

[14] J. K. Kuruvilla, A. Seth, J. Duttagupta, S. Sharma, and A. Jaiswal, Structural design and analysis of 6-DOF cylindrical robotic 

manipulators for automated agriculture. In Precision Agriculture for Sustainability. USA: Apple Academic Press, 2024.  

[15] D. Zhang, Y. Xu, Z. Hou, J. Yao, and Y. Zhao, "Optimal design and kinematics analysis of 5-dof hybrid serial-parallel 

manipulator," Transactions of the Chinese Society of Agricultural Engineering, vol. 32, no. 24, pp. 69-76, 2016.  

[16] A. Amir, A. Verma, A. Goswami, A. Kabra, S. Nakhye, and S. Chaudhary, "Design and analysis of strawberry-picking 

industrial Robotic Arm," presented at the 2022 IEEE Bombay Section Signature Conference (IBSSC), IEEE, 2022.  

[17] D. Bauer, C. Bauer, A. Lakshmipathy, and N. Pollard, Fully printable low-cost dexterous soft robotic manipulators for 

agriculture. In AI for Agriculture and Food Systems. United Kingdom: Springer Nature, 2022.  

[18] F. Wang and Y. Fan, "Structural design and analysis of a picking robot arm using parallel grippers," Advances in 

Mechanical Engineering, vol. 16, no. 12, p. 16878132241304610, 2024.  https://doi.org/10.1177/16878132241304610 

[19] Z. Wang, L. Gong, Q. Chen, Y. Li, C. Liu, and Y. Huang, "Rapid developing the simulation and control systems for a 

multifunctional autonomous agricultural robot with ROS," in Intelligent Robotics and Applications: 9th International 

Conference, ICIRA 2016, Tokyo, Japan, August 22-24, 2016, Proceedings, Part I 9, Springer International Publishing, 2016, pp. 

26-39.  

[20] D. Sepúlveda, R. Fernández, E. Navas, P. González-de-Santos, and M. Armada, "ROS framework for perception and 

dual-arm manipulation in unstructured environments," presented at the Robot 2019: Fourth Iberian Robotics 

Conference: Advances in Robotics, Springer International Publishing, 2019.  

[21] W. Chen and F. Liu, "Design of a grasping robot control system using kinematics model," Journal of Physics: Conference 

Series, vol. 2083, no. 2, p. 022031, 2021. https://doi.org/10.1088/1742-6596/2083/2/022031 

[22] Dassault Systèmes, Solidworks (Version 2022). France: Dassault Systèmes, 2023.  

[23] Open Robotics, ROS 2 humble Hawksbill (Version 2). United States: Open Robotics, 2022.  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Views and opinions expressed in this article are the views and opinions of the author(s), Current Research in Agricultural Sciences shall not be responsible or 
answerable for any loss, damage or liability etc. caused in relation to/arising out of the use of the content. 

 

https://doi.org/10.3389/fmech.2025.1543967
https://doi.org/10.1177/16878132241304610
https://doi.org/10.1088/1742-6596/2083/2/022031

