Publication
Online and Offline Analysis of Agricultural Machine Fleet data – Big Data Architecture Design and Evaluation
Zubair Jaleel
Mastersthesis, Technische Universität Kaiserslautern, 5/2018.
Abstract
CLAAS is an agricultural machinery manufacturing company which manufactures Tractors, Harvesters, Balers, etc. These machines carry out several agricultural activities to increase productivity and efficiency in the whole process chain. Our focus here is on the machines involved in the harvesting process. Harvesters along with Tractors are used for harvesting the agricultural crop grown for animal fodder, human consumption and biogas generation. They are manufactured with onboard sensors and devices which can record the data about the machine and the agricultural activity it does during the farming process. CLAAS collects this raw data for further processing to get useful information out of it. Analyzing this data will help us monitor and improve the farming activity and the machines involved in it. The current architecture in place at CLAAS supports only monitoring and storing of the harvest data but it does not support the varied analysis tasks which we need to perform on this data. It lacks the infrastructural supports to carry out these analysis tasks. We need an architecture with right tools and technology which can accomplish the required analysis tasks. In this Thesis, various data analysis tasks which are of business interest are gathered and the limitations of the current architecture to implement those analysis tasks are studied. Requirements for a new architecture is elicited and then come up with new architecture that meet those requirements. The capabilities and limitations of the new architecture are discussed.