Related Experiment Video
Updated: May 5, 2026
06:22
A Precise and Autonomous System for the Detection of Insect Emergence Patterns
Published on: January 9, 2019
6.2K
An AI-powered smart Agribot for detecting locusts in farmlands using IoT and deep learning
Mana Saleh Al Reshan1,2, Wahidur Rahman3,4, Shisir Mia3
1Department of Information System, College of Computer Science and Information Systems, Najran University, Najran, 61441, Saudi Arabia.
Scientific Reports
|November 13, 2025
Summary
This study developed an Agriculture Robot (Agribot) using Internet of Things (IoT) and Machine Learning (ML) for effective locust detection. The Agribot achieved 99.51% accuracy, demonstrating its feasibility for real-time agricultural pest management.
Area of Science:
- Agricultural Engineering
- Robotics
- Computer Science
Background:
- Locust infestations pose a significant threat to global agricultural production.
- Traditional methods for locust control are often insufficient to prevent widespread damage.
- Advanced technologies offer potential solutions for early detection and management of locust swarms.
Purpose of the Study:
- To develop an intelligent Agriculture Robot (Agribot) for real-time locust detection in agricultural fields.
- To integrate Internet of Things (IoT), Machine Learning (ML), and Deep Learning (DL) for enhanced pest detection capabilities.
- To evaluate the performance, speed, and usability of the developed Agribot system.
Main Methods:
- An IoT framework with sensors, an Android application, and a cloud server was implemented for automation.
- Pre-trained Convolutional Neural Network (CNN) models (VGG19) were utilized with ML classifiers (Logistic Regression) and feature selectors (SVC).
- A nature-inspired algorithm (Artificial Bee Colony - ABC) was incorporated into the ML/DL architecture.
Main Results:
- The Agribot achieved a maximum locust detection accuracy of 99.51% using the VGG19 CNN model with Logistic Regression and SVC.
- The Agribot demonstrated efficient operation with live video streaming at a maximum speed of 0.3048 m/s.
- The system achieved a System Usability Scale (SUS) score of 86%, indicating high user satisfaction.
Conclusions:
- The developed Agribot system shows strong feasibility for real-time locust detection and agricultural automation.
- The integration of IoT, ML, and DL provides an effective solution for managing agricultural pests.
- Further research can address identified limitations to optimize the system for broader application.
