Tag Archives: data mining

41-53 J. Kaivosoja and R. Linkolehto
Spatial overlapping in crop farming works
Abstract |
Full text PDF (791 kB)

Spatial overlapping in crop farming works

J. Kaivosoja* and R. Linkolehto

Natural Resources Institute Finland (LUKE), Green Technology, Vakolantie 55, FI 03400 Vihti, Finland
*Correspondence: jere.kaivosoja@luke.fi


A good driving accuracy and a proper machine input control are essential in sustainable farming. The goal is to work on the field exactly according to the plan, for example spraying a certain amount of fungicides evenly to the field. However, without modern assisting systems the farmers tend to overlap their driving lines. So far there have not been quantitative tests to present how mark able this overlapping is in real conditions. To solve this, we collected data from regular farming practices during four years in 17 different fields by recording GNSS (Global navigation satellite system) positions and a relative working status of the implement. We developed data mining methods of finding out the average overwork percentages of regular crop farming practices within different complete field plots. Based on the cumulative work distance, we measured the minimum overlapping percentage of different field works. The average minimum overlapping percentages for different machinery were: sprayer 15.7%, combine driller 7.7%, combine harvester 1.7%, spin disk fertilizer 9.5%, cultivator 19% and roller 59%. To understand reasons for great deviation between similar works, we determined different overlapping components for the spraying work: 2/3 of the spraying overlap was because of the driving line inaccuracies while the remaining 1/3 happens in the headland turns. This detected overlapping leads to the over consumption of pesticides, seeds, fertilizers, fuel and time but it can be minimized by applying accurate steering assistance and by adapting automatic section controls.

Key words:

, , , ,