Kubota is a pioneer among agricultural equipment manufacturers in Japan in spearheading research on smart agriculture, and it aims to attract more interest in the "reduction of labor requirements through the automatization and unmanned operation of agricultural machinery" as well as "data-based high-precision agriculture." In particular, Kubota will be implementing Nvidia's end-to-end AI platform to facilitate the reduction of labor requirements through the automatization and unmanned operation of agricultural machinery.
Hitherto, Kubota has been attempting to develop unmanned agricultural machinery that operates autonomously through the installation of computer vision. However, in order to create a completely unmanned agricultural machinery that can determine the appropriate form of agricultural operation from data such as weather and cultivation conditions and proceed to execute operations such as the harvesting of crops in a timely manner, the company will be carrying out R&D by utilizing the embedded AI platform "Nvidia Jetson" that is used for Nvidia's edge devices.
It will also be carrying out R&D by implementing the "Nvidia DGX AI system" as its AI training infrastructure. The DGX system supports NGC, a deep learning software hub optimized for GPUs, allowing developers to easily design, train, experiment with, and deploy AI models. Kubota is currently working on the development of an AI model capable of high-performance inference by utilizing Nvidia TensorRT, a key library for facilitating optimal AI inference performance.