Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay
Este es un artículo de acceso abierto distribuido bajo los términos de una licencia de uso y distribución CC BY-NC 4.0. Para ver
una copia de esta licencia visite http://creativecommons.org/licenses/by-nc/4.0/ 32
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay
Este es un artículo de acceso abierto distribuido bajo los términos de una licencia de uso y distribución CC BY-NC 4.0. Para ver
una copia de esta licencia visite http://creativecommons.org/licenses/by-nc/4.0/
Design and Development of IoT-based Harvesting Robo-Vec
Diseño y desarrollo de Robo-Vec de recolección basado en IoT
Projeto e desenvolvimento de colheita Robo-Vec baseada em IoT
Sadiq Ur Rehman
1
(*), Yasmin Abdul Wahab
2
Recibido: 13/08/2024 Aceptado: 30/10/2024
Summary. - This article presents "Harvesting Robo-Vec" an IoT-based autonomous harvesting robot designed to
enhance agricultural efficiency and precision. Integrating IoT technology with traditional methods, the robot automates
tasks and offers real-time monitoring and control. It navigates crop fields autonomously, detects ripe produce using
advanced sensing and imaging technologies, and performs precise harvesting maneuvers. Harvesting Robo-Vec
features an IoT communication module for seamless connectivity with a centralized control system, enabling remote
management of multiple robots. The paper outlines the robot's architecture, including its mechanical structure, sensors,
control algorithms, and communication infrastructure, along with safety, power management, and robustness
considerations. Iterative design, prototyping, and testing refined the robot's performance. Experimental results show
that Harvesting Robo-Vec improves efficiency, reduces labor costs, and enhances productivity compared to manual
methods. This study underscores the potential of IoT-based robots in agriculture, contributing to precision farming and
autonomous robotics research.
Keywords: IoT, Harvesting Robot, Computer Vision, Agriculture, Automation.
1
Ph.D., Assistant Professor, FEST, Hamdard University (Pakistan), sadiq.rehman@hamdard.edu.pk,
ORCID iD: https://orcid.org/0000-0002-6308-450X
2
Ph.D., Assistant Professor, NANOCAT, Universiti Malaya (Malaysia), yasminaw@um.edu.my,
ORCID iD: https://orcid.org/0000-0002-1681-2201
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 33
Resumen. - Este artículo presenta "Harvesting Robo-Vec", un robot de recolección autónomo basado en IoT diseñado
para mejorar la eficiencia y precisión agrícola. Al integrar la tecnología IoT con métodos tradicionales, el robot
automatiza tareas y ofrece monitoreo y control en tiempo real. Navega por los campos de cultivo de forma autónoma,
detecta productos maduros utilizando tecnologías avanzadas de detección e imágenes y realiza maniobras de cosecha
precisas. Harvesting Robo-Vec cuenta con un módulo de comunicación IoT para una conectividad perfecta con un
sistema de control centralizado, lo que permite la gestión remota de múltiples robots. El documento describe la
arquitectura del robot, incluida su estructura mecánica, sensores, algoritmos de control e infraestructura de
comunicación, junto con consideraciones de seguridad, gestión de energía y robustez. El diseño iterativo, la creación
de prototipos y las pruebas refinaron el rendimiento del robot. Los resultados experimentales muestran que Harvesting
Robo-Vec mejora la eficiencia, reduce los costos de mano de obra y mejora la productividad en comparación con los
métodos manuales. Este estudio subraya el potencial de los robots basados en IoT en la agricultura, contribuyendo a
la investigación en agricultura de precisión y robótica autónoma.
Palabras clave: IoT, Robot cosechador, Visión por computadora, Agricultura, Automatización.
Resumo. - Este artigo apresenta o "Harvesting Robo-Vec", um robô de colheita autônomo baseado em IoT projetado
para aumentar a eficiência e a precisão agrícola. Integrando a tecnologia IoT com métodos tradicionais, o robô
automatiza tarefas e oferece monitoramento e controle em tempo real. Ele navega pelos campos de cultivo de forma
autônoma, detecta produtos maduros usando tecnologias avançadas de detecção e imagem e realiza manobras de
colheita precisas. O Harvesting Robo-Vec apresenta um módulo de comunicação IoT para conectividade perfeita com
um sistema de controle centralizado, permitindo o gerenciamento remoto de vários robôs. O artigo descreve a
arquitetura do robô, incluindo sua estrutura mecânica, sensores, algoritmos de controle e infraestrutura de
comunicação, juntamente com considerações de segurança, gerenciamento de energia e robustez. O design iterativo,
a prototipagem e os testes refinaram o desempenho do robô. Os resultados experimentais mostram que a colheita
Robo-Vec melhora a eficiência, reduz os custos de mão-de-obra e aumenta a produtividade em comparação com os
métodos manuais. Este estudo ressalta o potencial dos robôs baseados em IoT na agricultura, contribuindo para a
agricultura de precisão e a pesquisa em robótica autônoma.
Palavras-chave: IoT, Robô de colheita, Visão computacional, Agricultura, Automação.
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 34
1. Introduction. - Agriculture uses traditional manual harvesting methods that are labor-intensive, time-consuming,
and prone to human error. These methods find it difficult to meet the expanding demands of a globalizing population
[1-2]. In addition, the problem is exacerbated by the dire consequences resulting from a shortage of personnel in
numerous sectors. Creative solutions that may automate and optimize the harvesting process are therefore desperately
needed to boost output, reduce costs, and improve crop quality [3].
The work on Internet of Things (IoT) [4-5] based smart agriculture monitoring systems in [6] offers a noteworthy
breakthrough by employing multiple algorithms to identify, measure, and evaluate vegetable development. Integrating
computer vision techniques and machine learning [7], these systems achieve over 90% accuracy, particularly focusing
on tomato cultivation. The development of autonomous smart agriculture robots, such as the Agri-Bot in [8], further
revolutionizes farming by performing labor-intensive tasks like planting, plowing, fertilizing, and harvesting,
leveraging Arduino UNOs and NodeMCUs for seamless automation. In [9], intelligent tomato-picking robots
demonstrate considerable improvements in agricultural efficiency, employing precise grasping mechanisms, enhanced
color segmentation, and advanced vision positioning to achieve an 83.9% success rate. Similarly, mechanical
harvesting robots for fresh-eating tomatoes in [10], equipped with stereo visual units, end-effectors, and rail-based
carriers, achieve an 83% success rate, significantly enhancing productivity and reducing labor costs.
Apple harvesting robots [11], featuring geometrically optimized manipulators, pneumatic grippers, and vision-based
recognition systems, further illustrate the potential of robotic technology in agriculture, successfully harvesting apples
with a 77% success rate. The integration of IoT and wireless sensors [12] in smart agriculture marks a transformative
shift from statistical to quantitative methods [13], exploring the potential of UAVs [14], precision farming, and wireless
sensors while discussing the benefits and challenges these technologies present. In [15] the design of greenhouse
tomato-picking robot chassis showcases advancements in precise positioning and cruising capabilities through
kinematic models [16], simulations, and physical testing, ultimately increasing the efficiency of greenhouse harvesting.
These studies highlight the advancements in IoT-based agriculture monitoring systems and robotic harvesting
technologies, underscoring their potential to revolutionize modern farming practices. Considering the research gaps,
we aim to create an innovative solution for the agricultural sector by developing an autonomous harvesting robot
named "Harvesting Robo-Vec" that leverages IoT technology. The objectives of the proposed system include:
Develop a robust and efficient design for the harvesting robot, equipped with necessary sensors, actuators,
and control systems to perform autonomous operations in the field.
Incorporate IoT modules to enable real-time monitoring, data collection, and communication with a
centralized control system.
Make accurate harvesting decisions with minimal crop losses by effective detection and recognition of ripe
fruits and vegetables using computer vision algorithms and proximity sensors.
Through this initiative, we seek to revolutionize the agriculture sector with a cost-effective, scalable solution, leading
to increased output, reduced dependence on labor-intensive farming practices, and strides toward precision farming.
2. Proposed system model.- The Harvesting Robo-Vec is an innovative harvesting robot that automates the process
of picking tomatoes. It integrates multiple hardware and software (See Figure I) components to make the field more
efficient and accurate.
Figure I. System Block Diagram
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 35
2.1 Hardware Components
2.1.1 Microcontrollers. - The ESP32 microcontroller [17] is a low-power microcontroller, both Wi-Fi and Bluetooth
capable. It is responsible for the communication line between different sensors and the robotic arm. It manages to
aggregate data retrieval, and actuation as a whole well. The main processing unit is the Raspberry Pi Model 3B+ [18],
which works in conjunction with the ESP32. This mini-computer has a complete Linux OS and handles complicated
functionalities like image processing and computer vision algorithms
2.1.2 Sensors. - Sensors in use provide the data about the environment to the system. The Pi Camera [19] takes high-
resolution images, and video of the crops, which is vital for processing visual information to identify ripe tomatoes.
Ultrasonic sensors [20] send out sound waves and measure the distance to nearby obstacles using the time-of-flight of
the received echo. Additionally, proximity sensors mounted on the robotic arm ensure that the distance to the tomatoes
is accurately gauged for effective harvesting.
2.1.3 Actuators. - The robotic arm is powered by high-torque servomotors, which enable precise control over its
movements, such as rotating and gripping. Motor driver modules interface with the microcontroller and servomotors,
allowing for control over speed and direction.
2.1.4 Power Management. - The robot’s operations are sustained by a battery management system (BMS) that
monitors and regulates the rechargeable batteries, ensuring they are charged safely and used efficiently. Buck converters
help manage voltage levels for different components, maintaining a stable power supply. Table I. presents the battery
configuration for the Harvesting Robo-Vec.
Parameter
Value
Number of Cells
5
Nominal Voltage per Cell
3.7 V
Total Voltage
18.5 V
Estimated Total Current
6.35 A
Operational Duration
4 hours
Battery Capacity
25.4 Ah
Total Energy Capacity
469.9 Wh
Table I. Battery configuration
Additionally, Table II presents an analysis of energy consumption along with suggested improvements and Figure II
is regarding the hardware used in the proposed model.
Energy Consumption
Factor
Current Issues
Expected Amperes (A)
Suggested Improvements
Microcontrollers
High power draw
from ESP32 and
Raspberry Pi.
ESP32: 0.15 A
Optimize algorithms for energy
efficiency.
Consider edge computing to
reduce Raspberry Pi load.
Raspberry Pi: 1.2 A
Sensors
Continuous
operation leads to
increased power
usage.
Pi Camera: 0.5 A
Implement smart sensor activation
(on-demand use).
Utilize low-power modes during
inactivity.
Ultrasonic Sensors: 0.1 A
each
Proximity Sensors: 0.1 A
each
Actuators
High energy
consumption for
servomotors.
2 A (per motor, typically)
Use variable torque control based
on task needs.
Explore energy recovery systems
(e.g., regenerative braking).
Table II. Energy consumption of Harvesting Robo-Vec
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 36
Figure II. Hardware used in the proposed model
2.2. Software Components
2.2.1 Operating System. - The Raspberry Pi operates on a Linux-based system, which provides a robust platform for
running applications and managing resources.
2.2.2 Programming Languages. - The software is developed using Python for applications running on the Raspberry
Pi, particularly for image processing tasks. The ESP32 microcontroller is programmed using C/C++, which is ideal for
handling low-level control and real-time sensor data processing.
2.2.3 Image Processing. - OpenCV [21] is a critical library used for computer vision tasks. It aids in tasks such as
color detection and image filtering, allowing the robot to identify ripe tomatoes based on their color and shape. The
system transforms images to the HSV color space for better color differentiation.
2.2.4 Control Algorithms. - The gripping and harvesting algorithm modulates the final gripping strength and
positioning given to feedback from the proximity sensors. This guarantees that tomatoes are taken carefully to protect
them from damage during harvesting. Using the ultrasonic sensors as input, the obstacle avoidance algorithm makes
sure the robot does not bump into obstacles while moving. Dijkstra's algorithm [22] calculates obstacles and determines
the optimal path to get through the crop field.
2.2.5 Communication Protocols. - The communication between Raspberry Pi and ESP32 is based on RESTFull API
[23] that allows command executing and data transmitting. This provides connectivity for remote control of the robot
and ensures seamless communication between all components
In Figure III, the flowchart shows the entire process of the proposed system.
Figure III. Process flowchart.
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 37
3. Testing Procedures and Experimental Setup.- Since the testing procedure and setup will be vital for assessing the
performance and functionality of Harvesting Robo Vec. The testing methods evaluate the performance parameters of
Harvesting Robo Vec in tomato identification and detection. The experimental frame for testing the performance of
the robot was established by simulating real agricultural environments incorporating a variety of factors. In the tomato
plantation, the cultivated areas consisting of plants were organized to replicate the arrangement and density of plants
in various fields, enabling a determination of robot efficiency to detect ripe tomatoes while navigating through diverse
arrangements. To ensure the efficient maneuvering of the robot, different obstetrical were designed to test the real
farming challenges which include soil types, and terrain variations. Moreover, the testing was performed concerning
various illuminations, considering the different times of the day and weather (e.g. sunny and cloudy) conditions.
Finally, the testing included levels for different soil types (e.g., clay, sandy, and loamy). This enabled us to compare
the collective effects of these factors on the robot's ability to harvest in terms of stability, traction, and efficiency.
Different testing scenarios were created to assess distinct operations of Harvesting Robo Vec with different situations
and are highlighted in Table III.
Description
Soil
Type
Crop
Variety
Lighting
Condition
Obstacle
Configuration
Standard Row Planting
Loamy
Ripe
Tomatoes
Clear,
Midday
Few Rocks
Cluster Planting with Intermixed Weeds
Sandy
Ripe
Tomatoes
Overcast
Wooden Fences
Row Planting with Uneven Growth
Clay
Ripe
Tomatoes
Clear, Early
Morning
Tall Grass
Standard Row Planting with Different
Tomato Varieties
Loamy
Cherry
Tomatoes
Clear, Late
Afternoon
Few Rocks
Mixed Crop Field (Tomatoes with Other
Vegetables)
Sandy
Mixed
Crops
Rainy
Bushes
Table III. Experimental Conditions and Setup
Key metrics analyzed included detection accuracy, which measures the percentage of correctly identified tomatoes,
providing insights into the reliability of the detection algorithms. Harvesting efficiency was also assessed, quantified
by the number of tomatoes harvested per minute, offering a clear indicator of the robot's productivity.
All sensor data, camera data, and performance logs were used to analyze how well the robot performed in the
agricultural environment. The use of detection accuracy, which explains the percentage of correctly identified
tomatoes, provides information about the reliability of the detection algorithms and thus, it will be one of the key
metrics analyzed in this work. We also evaluated harvesting efficiency (i.e., the number of harvested tomatoes per
minute), providing a direct measure of the productivity of the robot. Moreover, the time spent performing the harvesting
task was also recorded to assess the overall efficiency. Finally, the damage rate was measured as the percentage of
damaged tomatoes during harvesting, which is crucial for understanding the impact of the robot's operations on crop
quality as can be seen in Table IV.
Test
Condition
Detection Accuracy
(%)
Harvesting Efficiency (Tomatoes/Min)
Time Taken
(Min)
Damage Rate
(%)
Test 1
92
10
15
5
Test 2
85
8
20
10
Test 3
78
6
25
15
Test 4
90
9
18
6
Test 5
80
7
22
12
Table IV. Performance Metrics under Different Conditions
For this purpose, a comparative analysis between the proposed Harvesting Robo Vec and traditional manual harvesting
methods was carried out as shown in Table V which demonstrates the effectiveness and efficiency of the proposed
Harvesting Robo Vec over the traditional manual tomato harvesting method.
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 38
Method
Average Detection
Accuracy (%)
Average Harvesting Efficiency
(Tomatoes/Min)
Average Time
Taken (Min)
Average
Damage
Rate (%)
Manual
Harvesting
90
8
15
6
Robo Vec
Harvesting
92
10
12
4
Table V. Comparison of Harvesting Methods
4. Results and Analysis. - In the results and analysis of Harvesting Robo-Vec, a more complex analysis of
objectives was performed on the robot's live-operating performance against multiple metrics.
4.1 Tomato Detection Accuracy. - In regard to tomato detection accuracy results, the confusion matrix [24] (see
Table. VI), summarizes the performance of the algorithm in terms of true positive (TP), true negative (TN), false
positive (FP), and false negative (FN) detections. The confusion matrix shows the classification results and is used to
assess the accuracy of the tomato detection algorithm.
Predicted
Actual Tomato
Actual Not Tomato
Tomato
True Positive (TP)
False Positive (FP)
Not Tomato
False Negative (FN)
True Negative (TN)
Table VI. Confusion Matrix Key Term
To obtain the essential parameters, we used the following formulae:
Accuracy: (TP + TN) / (TP + TN + FP + FN) Eq (1)
Precision: TP / (TP + FP) Eq (2)
Recall: TP / (TP + FN) Eq (3)
By using data of TP = 120, TN = 700, FP = 150, and FN = 10, an accuracy of 83.67%, precision of 44.44%, and recall
of 92.31% was obtained. The result for the tomato detection with other objects placed and with only tomato(s) can be
seen in Figure IV.
Figure IV. Tomato detection.
4.2 Obstacle Avoidance Performance. - In terms of obstacle avoidance, the robot successfully navigated around 95%
of obstacles, with only 2 near misses and 3 collisions out of 100 encounters. This high success rate highlights the
effectiveness of the obstacle detection and avoidance systems, though further fine-tuning could enhance performance.
4.3 Robotic Arm Manipulation Efficiency. - The robotic arm's manipulation efficiency was impressive, achieving a
90% success rate in gripping attempts and an 85% success rate in harvesting tomatoes. These metrics underscore the
arm's reliability and effectiveness, with potential for further optimization in control algorithms and gripper design to
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 39
improve precision and reduce damage to the product. Figure V represents the gripping and dropping of tomato from
the robotic arm at the predefined location.
Figure V. Robotic Arm Manipulation.
The chassis of the vehicle, its charging port, and the resting position of the robotic arm placed in the final product
can be seen in Figure VI.
Figure VI. Robotic Arm Manipulation.
4.4 User Interface. - The prototype is controlled remotely using an Android application named "Harvesting Robot"
(see Figure VII). To operate it, the Android phone connects to the access point of the ESP-32 module named
"Harvesting Robot”.
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 40
Figure VII. Remote Control of Robo-Vec.
The User Interface for controlling and monitoring the Harvesting Robo-Vec is designed as an intuitive Android
application, enabling remote operation of the robot. Key features include straightforward control options that allow
users to remotely direct the robot’s movements and harvesting actions. The app provides manual control capabilities,
as well as options for pausing and resuming operations as needed. Additionally, users can adjust specific parameters,
such as speed and harvesting sensitivity, to tailor the robot’s performance to various field conditions or crop types. The
total cost analysis of IoT-based Harvesting Robo-Vec is presented in Table VII.
Component
Cost Estimate (Pkr)
Development and Prototyping
15,000
Microcontrollers (ESP32, Raspberry Pi)
37,500
Sensors (camera, ultrasonic, proximity)
2,600
Actuators (servomotors)
5,000
Battery (Li-ion system)
750
Chassis and Frame
5,000
Software Development
5,000
Contingencies
6,000
Total Initial Investment
76,850
Table VII. Cost analysis of IoT-based Harvesting Robo-Vec
5. Limitations and Failure Modes. - While the Harvesting Robo-Vec indeed works autonomously and can be a useful
tool in agricultural scenarios, it does have its limitations and possible problems that could arise with its operations.
We, therefore, explore these hurdles and propose solutions.
5.1 Obstacle Proximity Challenges. - The robot may have a problem identifying obstacles that are too close, which
can cause collisions, or the robot to unexpectedly stop. This can be compensated for by adding additional sensors and
moving their position
5.2 Detection Errors in Tomato Identification. - Tomato harvesting depends on precise detection with no leading
fault. Such reliability can also be achieved using advanced image processing and methods of machine learning.
5.3 Energy Consumption and Limited Battery Life. - Prolonged operation can lead to quick battery depletion,
especially under heavy workloads. Implementing energy-saving modes and optimizing operational paths can help
extend battery life.
5.4 Mechanical Wear and Tear. - If the robot has run for some time there can be wear in the robotic components that
can affect the performance. This problem can be minimized by making use of durable materials and also by following
regular maintenance schedules.
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 41
6. Ethical and Social Implications. - Harvesting Robo-Vec will raise real ethical and social considerations that need
to be addressed.
6.1 Impact on Jobs. - Automation will improve efficiency and help reduce service costs; It will also decrease the
traditional demand for farm workers. Conversely, this tech can also generate employment in terms of maintaining,
operating, and managing the robot systems, which can assist workers with the transition into new positions.
6.2 Environmental Considerations. - The Robo-Vec is designed to be energy-efficient, helping to lessen its
environmental impact through smart power management and low-energy components. Future models might even use
renewable energy sources, such as solar panels, to further decrease reliance on traditional power, supporting more
sustainable farming practices.
6.3 Workers Safety. - Proximity sensors and emergency stop buttons are also added to Robo-Vec to ensure the safety
of human workers around it. They even have sensors that are sensitive enough to detect anybody walking by, ultimately
stopping the robot from doing its task to avoid an accident
With the continuous expansion of agricultural robotics, it is necessary to consider these ethical and social dimensions
to ensure these advancements benefit society and the environment.
7. Conclusion and Future Work. - The proposed system model successfully achieved its goals and made significant
advancements in automated tomato harvesting. The meticulous design and integration of the robotic system's
components resulted in an effective and reliable outcome. The implementation of a Raspberry Pi-based color
recognition algorithm enabled accurate tomato detection, while the integration of the ESP32 microcontroller facilitated
seamless movement control, obstacle detection, and robotic arm manipulation. The findings indicate that the IoT-based
Harvesting Robo-Vec has the potential to revolutionize tomato harvesting by reducing manual labor, increasing
productivity, and ensuring consistent results through the successful integration of hardware, software algorithms, and
IoT capabilities. Future work could enhance the tomato detection system's accuracy and robustness using machine
learning techniques or advanced image processing algorithms and improve control through real-time data analytics
and remote monitoring via IoT connectivity. Future iterations might also incorporate renewable energy sources, such
as solar power, to enhance the system's sustainability and operational efficiency.
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 42
References
[1] Rehman, S.U., Zaidi, A.A., Wahab, Y.A., Arafat, M.Y. and Hatta, S.F.W.M., “Advanced Solar-Powered Seed
Sowing Machine with Precision Seeding and Smart Control Features”. IEEE Regional Symposium on Micro and
Nanoelectronics (RSM) pp. 134-137. IEEE. August 2022. https://doi.org/10.1109/RSM59033.2023.10326842
[2] Haider, W. and Rehman, A., "Knowledge-based Soil Classification Towards Relevant Crop Production". Int. J.
Adv. Comput. Sci. Appl., 10(12), pp.488-501. 2019. https://doi.org/10.14569/IJACSA.2019.0101266
[3] Haider, W., Rehman, A.U., Durrani, N.M. and Rehman, S.U., “A generic approach for wheat disease classification
and verification using expert opinion for knowledge-based decisions”. IEEE Access, 9, pp.31104-31129, 2021.
https://doi.org/ 10.1109/ACCESS.2021.3058582
[4] Aqeel-ur-Rehman, S.U.R., Khan, I.U., Moiz, M. and Hasan, S., “Security and privacy issues in IoT”. International
Journal of Communication Networks and Information Security (IJCNIS), 8(3), pp.147-157. 2016.
[5] Rehman, S.U., Mustafa, H., Shaikh, M.A. and Memon, S., “Towards Sustainable Energy Storage: A Low-Cost
IoT Solution for Real-time Monitoring of Lead-Acid Battery Health”. Memoria Investigaciones en Ingeniería, (26),
pp.202-212, 2024. https://doi.org/10.36561/ING.26.12
[6] Siddiquee, K.N.E.A., Islam, M.S., Singh, N., Gunjan, V.K., Yong, W.H., Huda, M.N. and Naik, D.B.,
“Development of Algorithms for an IoT‐Based Smart Agriculture Monitoring System”. Wireless Communications
and Mobile Computing, 2022(1), p.7372053. 2022.
[7] Attri, Ishana, Lalit Kumar Awasthi, and Teek Parval Sharma. "Machine learning in agriculture: a review of crop
management applications." Multimedia Tools and Applications 83, no. 5 (2024): 12875-12915.
https://doi.org/10.1155/2022/7372053
[8] Rai, H.M., Gupta, D., Mishra, S. and Sharma, H., “Agri-Bot: IoT Based Unmanned Smart Vehicle for Multiple
Agriculture Operation”. International Conference on Simulation, Automation & Smart Manufacturing (SASM), pp.
1-6. IEEE. August 2021. https://doi.org/10.1109/SASM51857.2021.9841182
[9] Feng, Q., Wang, X., Wang, G. and Li, Z., “Design and test of tomatoes harvesting robot”. IEEE international
conference on information and automation, pp. 949-952. IEEE. August 2015.
https://doi.org/10.1109/ICInfA.2015.7279423
[10] Feng, Q., Zou, W., Fan, P., Zhang, C. and Wang, X., “Design and test of robotic harvesting system for cherry
tomato”. International Journal of Agricultural and Biological Engineering, 11(1), pp.96-100, 2018.
https://doi.org/10.25165/j.ijabe.20181101.2853
[11] De-An, Z., Jidong, L., Wei, J., Ying, Z. and Yu, C., “Design and control of an apple harvesting robot”. Biosystems
engineering, 110(2), pp.112-122, 2011. https://doi.org/10.1016/j.biosystemseng.2011.07.005
[12] Rehman, A.U., Rehman, S.U. and Raheem, H., “Sinkhole attacks in wireless sensor networks: A survey”.
Wireless Personal Communications, 106, pp.2291-2313, 2019. https://doi.org/10.1007/s11277-018-6040-7
[13] Ayaz, M., Ammad-Uddin, M., Sharif, Z., Mansour, A. and Aggoune, E.H.M., “Internet-of-Things (IoT)-based
smart agriculture: Toward making the fields talk”. IEEE access, 7, pp.129551-129583, 2019.
https://doi.org/10.1109/ACCESS.2019.2932609
[14] Pandey, G.K., Gurjar, D.S., Yadav, S., Jiang, Y. and Yuen, C. “UAV-Assisted Communications With RF Energy
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 43
Harvesting: A Comprehensive Survey”. IEEE Communications Surveys & Tutorials, 2024.
https://doi.org/10.1109/COMST.2024.3425597
[15] Su, L., Liu, R., Liu, K., Li, K., Liu, L. and Shi, Y., “Greenhouse tomato picking robot chassis”. Agriculture,
13(3), p.532, 2023. https://doi.org/10.3390/agriculture13030532
[16] Russo, M., Zhang, D., Liu, X.J. and Xie, Z., “A review of parallel kinematic machine tools: Design, modeling,
and applications”. International Journal of Machine Tools and Manufacture, p.104118, 2024.
https://doi.org/10.1016/j.ijmachtools.2024.104118
[17] El-Khozondar, H.J., Mtair, S.Y., Qoffa, K.O., Qasem, O.I., Munyarawi, A.H., Nassar, Y.F., Bayoumi, E.H. and
Abd El, A.A.E.B., “A smart energy monitoring system using ESP32 microcontroller.” e-Prime-Advances in Electrical
Engineering, Electronics and Energy, 9, p.100666, 2024. https://doi.org/10.1016/j.prime.2024.100666
[18] Shakir, M., Karim, S., Memon, S., Rehman, S.U. and Mustafa, H., “An improvement in IoT-based smart trash
management system using Raspberry Pi”. International Journal of Computational Vision and Robotics, 14(2),
pp.191-201, 2024. https://doi.org/10.1504/IJCVR.2024.136997
[19] Groffen, J. and Hoskin, C.J., “A portable Raspberry Pi‐based camera set‐up to record behaviours of frogs and
other small animals under artificial or natural shelters in remote locations”. Ecology and Evolution, 14(3), p.e10877,
2024. https://doi.org/10.1002/ece3.10877
[20] Khaleel, H.Z., Ahmed, A.K., Al-Obaidi, A.S.M., Luckyardi, S., Al Husaeni, D.F., Mahmod, R.A. and Humaidi,
A.J., “Measurement enhancement of ultrasonic sensor using pelican optimization algorithm for robotic application”.
Indonesian Journal of Science and Technology, 9(1), pp.145-162, 2024. https://doi.org/10.17509/ijost.v9i1.64843
[21] Zhang, Y. and Xu, D., “Design of an intelligent medication delivery robot based on CNN and OpenCV”. Ninth
International Symposium on Sensors, Mechatronics, and Automation System (ISSMAS 2023), vol. 12981, pp. 1451-
1458. SPIE, March, 2024. https://doi.org/10.1117/12.3014833
[22] Maristany de las Casas, P., Kraus, L., Sedeño‐Noda, A. and Borndörfer, R., “Targeted multiobjective Dijkstra
algorithm”. Networks, 82(3), pp.277-298, 2023. https://doi.org/10.1002/net.22174
[23] Ehsan, A., Abuhaliqa, M.A.M., Catal, C. and Mishra, D., “RESTful API testing methodologies: Rationale,
challenges, and solution directions”. Applied Sciences, 12(9), p.4369, 2022. https://doi.org/10.3390/app12094369
[24] Amperawan, A., Andika, D., Anisah, M., Rasyad, S. and Handayani, P., “Confusion Matrix Using Yolo V3-
Tiny on Quadruped Robot Based Raspberry PI 3B”, 7th FIRST 2023 International Conference on Global Innovations
(FIRST-ESCSI 2023),(pp. 549-562. Atlantis Press, February, 2024. https://doi.org/10.2991/978-94-6463-386-3_56
S. Ur Rehman, Y. Abdul Wahab
Memoria Investigaciones en Ingeniería, núm. 28 (2025). pp. 32-44
https://doi.org/10.36561/ING.28.4
ISSN 2301-1092 • ISSN (en línea) 2301-1106 – Universidad de Montevideo, Uruguay 44
Author contribution:
1. Conception and design of the study
2. Data acquisition
3. Data analysis
4. Discussion of the results
5. Writing of the manuscript
6. Approval of the last version of the manuscript
SUR has contributed to: 1, 2, 3, 4, 5 and 6.
TAW has contributed to: 1, 2, 3, 4, 5 and 6.
Acceptance Note: This article was approved by the journal editors Dr. Rafael Sotelo and Mag. Ing. Fernando A.
Hernández Gobertti.