Decision support in a smart greenhouse using large language model with retrieval augmented generation

Abstract

The paper discusses the application of large language models for the development of decision support systems, particularly in the agro-industry. The thesis that decision support is necessary in the sphere of agriculture is confirmed. In order to optimize the activity of smart greenhouses it is reasonable to implement information systems and technologies using large language models. The architecture of decision support system with Retrieval Augmented Generation technology for controlling disease incidence of plants cultivated is proposed. Experiments with the implemented decision support system have been carried out. It is concluded that the application of large language models in the operation of smart greenhouses is promising for use in agriculture in general.