Conceptual development of a precision-based robot design for Green Field agricultural applications using IoT, Artificial Intelligence & Machine Learning techniques

Abstract

For many years, agriculture has played a vital role in the Indian economy, contributing significantly to the nation's Gross Domestic Product (GDP), with the agricultural sector accounting for 18% of India's GDP. Robotics and the Internet of Things (IoT) present a promising solution for advancing precision agriculture. Particularly, the selective harvesting of crops such as tomatoes and apples demands more precise and efficient methods. Conventional farming involves labors individually handpicking the ripened fruits which requires immense manpower to perform selective farming. We propose a robot that assists farmers in various labour-intensive tasks such as selective crop harvesting, qualitative segregation and also concurrently provide information of crop health, soil nutritional status and crop shelf-life detection. The collected information is analysed, processed and sent to the farmer via android application. Later, quality of the particular harvested crop is inspected by checking the weight, color, health to grade them. The graded fruit is then transferred to the assigned container. Spoiled or over ripened fruits/vegetables would be plucked and dropped so that it would not affect the growth of the plant. Proposed robot would have a harvesting arm which reaches the fruit/vegetable to pluck it from the plant or a tree which would later transfer the fruit to respective container. User would be able to assign a specific task to the device according to his requirement with the help of the app. Selective harvesting, segregation and crop health monitoring requires image processing through camera. It also detects diseases using image processing. We are doing the pre-processing and image segmentation. The guided, supervised, and advanced machine learning technique known as k-nearest neighbors (KNN) is employed to develop solutions for both classification and regression problems. Agriculture automation is a major concern and a hot topic in all countries. Given how quickly the world's population is expanding, there is an urgent demand for food. Farmers must use harmful pesticides more frequently because their current approaches cannot keep up with the rising demand. The soil is harmed by this. The land remains unproductive and barren as a result, which has a substantial impact on agricultural practises. IOT, wireless connectivity, AI & ML and deep learning are just a few of the automation methods that can be used. This provides the farmer with effective data that plays a vital role in selection of pesticides, herbicides and right kind of fertilizers required for increased yield.