The paper discusses creation of a digital twin (DT) of plant for an intelligent cyber-physical system for managing precision farming. A new approach to formalization of DT knowledge is proposed to form expert knowledge within the subject area based on the ontological specification of stages of plant growth and development and multi-agent technology for creating stage agents and coordinated dynamic recalculation of stage duration and yield forecast based on events in the environment. The paper proposes a method for calculating the forecast for duration of plant development stages and yield based on expert knowledge. A “tube” model of the range of changes in parameters of plant development for each stage has been developed. The paper also introduces a method for calculating the yield forecast, as well as the dates of beginning and end for each plant development stage within the “tube” during their normal development and in case of critical situations, for example, frost or drought. Ontology of plant development is constructed for implementation of the “tube” model of environmental parameters, which is converted into a digital form within the ontology editor, available for use by agents. The paper describes the structure and functions of a smart plant DT, built on the basis of a knowledge base and a module for multi-agent planning of plant development stages (for example, wheat), integrated with external weather forecast and fact services. A brief description of the created prototype of the intelligent plant DT system in Java is given. Using the system, agronomists can create their own knowledge bases and DTs of the cultivated plants for each field or even field section. The system will be useful in modern crop production for precision farming, not only “place-wise” but also “time-wise”, i.e. in terms of the best time for performing agrotechnical operations.
The paper presents precise agriculture as a complex adaptive system with high level of uncertainty and dynamics, in which knowledge is forming experimentally and in a very enterprise-specific way. Is there any opportunity to learn the best practices from advanced precise farmers, transfer their knowledge and support everyday decision making for regular farmers? The concept of Smart Farming as an augmented AI solution for precise agriculture is proposed. The solution is designed as a digital ecosystem (system of systems) of smart services, where each service, in its turn, is an autonomous AI system. The paper also discusses functionality of smart services for precise agriculture and the service-oriented architecture of the solution with p2p interaction of services. Ontology-driven knowledge base and multi-agent technology are considered as the key technologies of the solution. The virtual "round table" for coordinated decision making of smart services is introduced. Finally, the paper presents results of the first applications, as well as the future steps and expected results.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.