Ikka: An Intelligence garlic farm robot with AI-enhanced pest detection, automated pesticide application, and integrated weather forecasting capabilities

ABSTRACT: STATEMENT OF THE PROBLEM: This study aims to develop an Intelligent Garlic Farm Robot, named Ikka, equipped with AI-Enhanced Pest Detection, Automated Pesticide Application, and Integrated Weather Forecasting Capabilities. The purpose is to assist farmers in pest monitoring and pesticide application, and to improve the efficiency of the farming process. To achieve this, the researchers aim to answer the following research problems and questions: (1) the value of environmental monitoring in implementing Ikka, (2) the accuracy satisfaction of pest detection and pesticide application of Ikka, (3) the willingness of garlic farmers to utilize Ikka in garlic cultivation, and (4) the technological acceptance of utilize Ikka in garlic cultivation. To accomplish this, the specific objectives are to integrate an object detection system for pests and insects, develop web applications for weather forecasting and for users to provide commands to Ikka, and determine the technological acceptance and usage of Ikka. RESEARCH METHODOLOGY: This research is applied and focuses on developing a product and services that meet the needs of a specific customer segment. It’s conducted to Ilocos Norte, Philippines, targeting garlic farmers. The Iterative and Incremental Development model (IID) is used, which has five stages. To collect data, a structured survey questionnaire with 5-point Likert Scale response method and a feedback survey questionnaire with a 5-point Likert Scale response method will be used. Descriptive statistics will be utilized to analyze the data. The Ikka prototype will be presented to garlic farmers, and their feedback will be studied to determine the effectiveness of Ikka in addressing common problems faced by garlic farmers in the area. It will also be tested to determine the performance of the AI using the confusion matrix. The test consists of 6 healthy plants and 4 infected plants. Ikka will have to automatically scan each plant for 5 trials to be able to get an average accuracy, precision, recall, sensitivity, and f1-score of the AI. SUMMARY OF FINDINGS: The research investigates the attitudes and perceptions of garlic farmers towards a semi-autonomous farming robot. Ikka, with a pest and environment monitoring system. The study found that the most significant problem for farmers is the impact of pests and plant diseases on their livelihood, and the inability to predict weather accurately. The majority of respondents are open to using Ikka, trusting it to take care of their farm, and investing in it. They consider real-time monitoring of environmental conditions and the use of technology for pest detection important. A monitoring system for temperature, humidity, and wind information was rated as the most important feature. An AI model trained to detect pests and diseases achieved satisfactory accuracy rates, and respondents perceived Ikka as useful in their farming operations, with case of use positively perceived. The performance of the model was determined using the confusion matrix with the average of 84% for accuracy, 93.81% for precision, 80% for recall, 90% for specificity, and 85.31% for the f1-score. Farmers perceive Ikka as a technology that can reduce labor costs, make farming more efficient, reduce pesticide use, and have a positive impact on the environment . CONCLUSION: Garlic farmers are facing various obstacles in their farming journey such as pests, diseases, weather uncertainty, and extensive labor requirements. However, they are willing to embrace new technology and are open to using Ikka, a semi-autonomous farming robot that comes with a pest and environment monitoring system. According to respondents, real-time monitoring of environmental conditions and pest detection technology are valuable features. Ikka is perceived to be beneficial in reducing labor costs, optimizing efficiency, minimizing pesticide use, and having a positive environmental impact. It also highlights the satisfactory results of an AI model trained to detect pests and diseases in garlic plants. The performance of the model it also determined using the confusion matrix with the average results of 84% for accuracy, 93.81% for precision, 80% for recall, 90% for specificity, and 85.31% for the f1-score. Furthermore, respondents view Ikka as a useful tool in enhancing farming productivity, improving quality, speeding up tasks, and simplifying pest monitoring. RECOMMENDATION: In light of the aforementioned findings and conclusions, the following recommendations are presented: 1. Conduct further research on the impact of pests and plant diseases on garlic farming, with a focus on identifying effective strategies to mitigate their effects. 2. Conduct a feasibility study on the development and implementation of a semi-autonomous farming robot with a plant and environment monitoring system, taking into account the preferences and concerns of garlic farmers. 3. Develop a farming robot prototype with features rated highly by respondents, such as real-time monitoring of environmental conditions and image processing and object detection for pest detection. 4. Investigate the potential benefits and drawbacks of automated pesticide applications in garlic farming, taking into account the perceived low importance rating by respondents in the Importance of Farming Robot Features section. 5. Further refine and test the AI model for pest and disease detection in garlic plants to increase accuracy and identify potential improvements. 6. Conduct additional studies to gather more feedback on future research and explore opportunities to enhance its functionality and usability for garlic farmers.

Filipiniana