Drone-assisted climate-smart agriculture (DACSA): A spatially-based outcome prediction model as an initial approach to track yield changes in shallot planting areas

Abstract

The challenge of meeting the world's food demands while protecting the environment is an urgent concern. Leveraging Technology 4.0, precision agriculture management emerges as a promising solution to enhance efficiency and effectiveness. With this goal in mind, a research project was undertaken to develop a spatial model capable of providing real-time notifications on changes in plant yield and location. This map would enable farmers to monitor their fields more precisely and with greater detail. The research commenced by conducting direct soil measurements and capturing crop spectra using drones equipped with multispectral cameras. The collected data was then mosaicked, processed, and combined with crop yield data to create a set of samples. Machine learning algorithms were employed to make predictions, and the yield projections were integrated into spatial maps. These maps can be utilized for navigation and to track areas anticipated to experience yield changes. The result is a spatial map model that serves as a tracking and navigation system, empowering farmers to monitor crop yield changes effectively. This research is an integral part of the development of the Drone-Assisted Climate-Smart Agriculture (DACSA) system, which employs drones for tasks such as obtaining multispectral image data from onion plants, mapping, spraying, and fertilizing. Leveraging the input data spectrum, this spatial navigation map model is expected to contribute to the gradual implementation of precision agriculture, ensuring sustained productivity, and enabling the localization of crop issues in specific areas.