Agri-PAD: a scalable framework for smart agriculture

Abstract

More recently, big data tools and technologies have been applied in the agriculture sector leading to major benefits. Many frameworks have been proposed that employ big data technologies in the field of agriculture, however, such existing frameworks are focused on a particular aspect of agriculture and do not consider multiple stakeholders and applications. The objective of this research is to develop a holistic framework named Agri-PAD that encompasses almost all aspects of agriculture including crop selection, crop monitoring, soil monitoring, weather conditions, precision farming, and market demand. The Agri-PAD framework includes three major categories of machine learning based agriculture applications that is precision, recommendation, and enterprise applications. The Agri-PAD framework is capable of providing remote sensing of fields, precision farming, effective supply chain, and support informed decision making leading to enhanced productivity. To validate the efficacy of the proposed framework, the two most prominent agricultural applications, crop production forecasting and crop harvesting recommendation have been investigated and accuracy of 99% has been achieved. We believe that the Agri-PAD framework enables all stakeholders in the agriculture cycle to connect and apply big data analytics at every step leading to a more efficient and smarter agriculture ecosystem.