Design and Development of the Information System for Dairy Farms Indicators Prediction

There is an increase in the number of companies working in the agricultural industry in the recent years. However, the level of informatization of the industry is relatively low. In the dairy farming sub-industry, there is only a limited choice of information solutions – one group for the support of milking parlor and herd management (provided by machinery manufacturers), and solutions for accounting and tax reporting. The total milk yield value, which is an important indicator for farm managers to track is difficult to predict, as it depends on many factors. A number of scientific researches have been already conducted, which are devoted to the topic of different predictions of milk indicators. Many of them show positive successful results, but remains on paper and do not find implementation in real production systems. The design and development of the system which could provide predictions of the farm indicators could help such companies to make right decisions and thereby save their resources. The main goal of the work is to design such a system and develop some modules of it, which could predict valuable indicators for the operation activity of dairy farms. To reach the goal set it is necessary to complete the following objectives: review and analyze the current state of researches in the sphere of dairy farms indicators predictions, the existing information solutions; review the subject area practically on a operate dairy farm, identify indicators of a typical dairy farm, which could be predicted, develop the requirements for the system; design the proposed information system and develop some of the modules of it. As the result of the work all of the objectives have been completed, and the main goal has been achieved. The designed architecture and developed UI modules could be successfully used for a prediction system. However, the size and quality of existing data allows to make acceptable predictions for very short terms. Further researches could be devoted to the improvement of the model accuracy by adding additional parameters, potentially affecting yield values.