IBM-NASA Geo-Spatial Model Prithvi 100M for Predicting Agricultural Machinery Demand

The dynamic and region-centric requirements of agriculture are not met by conventional forecasting methods, which are primarily dependent on the timely availability and appropriate use of machinery. Agricultural productivity is primarily determined by these factors. In this paper, a novel methodology is proposed to estimate the demand for agricultural machinery needed for the first time by utilizing the IBM-NASA Geo-Spatial Model, Prithvi 100M. The Prithvi 100M model integrates cutting-edge machine learning algorithms with high-resolution satellite imagery to analyze and process soil conditions, crop types, and growth stages across geographical regions. We utilized data on machinery utilization and crop yield history to train the model, enabling it to ascertain the necessary type and quantity of agricultural machinery during various stages of a farming cycle. In order to verify that the model was functioning properly, it utilized actual data from certain plants in pilot regions to collect data on their machinery usage. The Prithvi 100M generates highly precise machinery predictions, reduced operating expenses, and enhanced business productivity. This study not only offers a comprehensive assessment of the models, but it also illustrates the extent to which resource allocation has been altered by this technology and has the potential to do so in the Agricultural Planning & Operations sector.