Masters Theses

Date of Award

5-2002

Degree Type

Thesis

Degree Name

Master of Science

Major

Biosystems Engineering Technology

Major Professor

William E. Hart

Committee Members

John B. Wilkerson, Robert S. Freeland

Abstract

The study analyzed operator feedback methods of four guidance systems used on agricultural spraying vehicles. One system was the traditional foam marker and the three others were GPS guidance systems, manufactured by Trimble, Satloc®, and Cultiva. Twenty-one drivers participated in this study and ran four tests over a 244 m course. An RTK GPS system provided horizontal center positioning data accurate to within a centimeter. Using the center positioning data, areas of overlap and skip were found using ArcView™ 3.2. In addition, positional error terms of maximum deviation, average maximum deviation, RMS, average, starting, and ending deviation were found. After completing the tests, drivers answered questions regarding the systems. The data set as a whole provided a means for determining differences between guidance systems, evaluating if effects of driver and system explained positional errors, and which systems did the drivers prefer.

Results of overlap and skip were analyzed to determine if differences existed between guidance systems, GPS guidance systems, speeds of high and low, and passes. Analyzing overlap for systems found that GPS systems differed from foam at the 0.05 alpha level, and that systems did not differ for skip. Mean values indicated that foam was higher than GPS guidance systems for both overlap and skip. GPS guidance systems did not differ for overlap and skip at the 0.05 alpha level. Mean values indicated that Cultiva had higher means followed by Satloc and Trimble respectively. Speeds of high and low did not differ for overlap and skip at the 0.05 alpha level. Plots of passes on a per driver and system basis did not indicate that a learning curve was present. Mean values of speed indicated that high speeds had a greater mean errors than slow speeds.

Results of positional errors were analyzed to study if driver and system affected positional errors and passes. Examination of the data revealed that when found significant, driver explained less than 20 % of the variation. Pass analysis on a per system basis revealed that a learning curve was not evident for Trimble, Satloc®, and Cultiva.

Examination of the survey data revealed that drivers did not prefer one system to the other. Of the twenty-one drivers, seven preferred Trimble’s lightbar, six preferred Satloc’s lightbar, and eight preferred Cultiva’s Marker™. However, drivers did prefer GPS guidance to foam. Fifteen drivers preferred GPS guidance to foam, while sixteen drivers felt that GPS guidance improved their accuracy.

Statistically, guidance systems only differed for overlap; however, mean values indicated that foam had higher mean error values associated with overlap and skip. This indicated that a driver’s accuracy improved when using GPS guidance. Statistically, GPS systems did not differ; however, mean values indicated that drivers were less accurate using Marker™, which revealed drivers were more accurate using lightbars as opposed to using a graphical display. Statistically, speeds of high and low did not influence accuracy; however, mean values indicated that high speeds had greater mean errors than slow speeds, and a learning curve was not distinguishable between passes.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS