Data Conceptual Model for Smart Poultry Farm Management System
Nikolajs Bumanis, Irina Arhipova, L. Paura, Gatis Vītols, Liga Jankovska
Abstract
Internet of things provides ways to obtain information about farm production; however, the more data there is, the harder it is to determine the correct way to process these data. The solutions on the market are aimed specifically at gathering data, rather than processing and analyzing them. Poultry farming is one of the fields, where the application of Industry 4.0 principles is becoming a necessity, especially in the case of greenhouse gas control. Minimum standards and rules in the poultry management are regulated by EU Directives and required parameters must be ensured at all times, therefore, real-time environmental monitoring must be performed. The only practically viable approach is to automate data gathering using appropriate sensors, which is only possible if a standardized data structure model is defined. In this research the Cyber-Physical Model is the proposed as a basis to development of smart poultry farm management system that is adjustable to particular production needs. Three main data groups are defined – necessary data to ensure requirements of EU Directives, farm monitoring data for business analysis and optimization, and additional data that can influence overall poultry farm management. The proposed data set was implemented into the existing infrastructure of a Baltic poultry farm. Multiple CO2 (carbon dioxide) and NH3 (ammonia) sensors were installed in order to gather data, measurements of which previously were taken manually by the farm’s staff. The solution was developed on centralized cloud-based data processing system, where MQTT Broker was used for security measures. The processed data is used by the decision support system with the aim to define optimal feeding and housing.