Proceeding of International Agriculture Innovation Conference

Abstract

In the paper, an design principles for developing of smart cyber-physical system for precise farming is given. It includes ontology-driven domain knowledge base in plants cultivation and multi-agent technology for plant growth planning and adaptive farm resource scheduling.

Smart cyber-physical system (CPS) includes digital twin of plant which triggers adaptive rescheduling of farm resources. Multi-agent technology provides agents for collective distributed decision making. To find the optimum solution in the decision making process, the agents are reaching competitive equilibriums on virtual market and are able to discover and solve conflicts by continuous, parallel and asynchrony, auction-like negotiations for real-time adjustment of precise farming operations.

Smart CPS with digital twin of plant is implemented through a web-interface and mobile application. In the article, some of the features and use-cases are shown. Smart CPS with digital twin key function is to simulate field processes and provide the best performance for the farm. It will give recommendations to the user, based on various sensors, indicators, facts and other information. Ontology-driven knowledge base provides the rules, restrictions and objects for decision making process. Accordingly to the developed interface, a user (an agronomist from a farm, for example) is able to guide all the process of plant cultivation digitally. The following services realizes the functionality of the system: crop rotation planning, operational planning, field monitoring, employees and machines load planning, finance charts. Knowledge bases for wheat and rice are under development.

Digital twin of plant will be implemented as a part of smart CPS. Digital twin of plant allows to simulate the growth process of real plant and predicts its’ effect for different agricultural events. Crop personification due to real-time control and management through CPS with plant digital twin is expected to significantly increase the value of cultivated products.

The developed web-interface of smart CPS has various functions. Thereby, relevant information related to the fields, crops, infrastructure is displayed. Field images, got with remote sensing, are processed to calculate the NDVI factor, to estimate the chemical components and to track the autonomous vehicles for precise farming. Business monitoring helps to pick all the data, related to different crops, on the same screen for better visibility. Basic economical forecasts could be another indicator for the agronomist to define the preferred scenario. Mobile application allows access to the system from the field and implements key functions to be accessible offline.

The system will help agronomists and farmers to make the next step in digital transformation of agriculture, maximize return of investments from precision farming, automate

decision-making processes for farmers and service companies and make their business smarter, more flexible, faster, easier and more cost-efficient, provide better productivity of plant cultivation and sustainability of agriculture under global climate changes.

Keywords

cyber-physical system, multi-agent technologies, precise farming, digital twin, knowledge base, ontology, plants cultivation, distributed decision making