Precision pesticide delivery based on a GIS map

A sprayer to deliver agrochemicals based on an aerial scan of a Nova Scotian blueberry field was developed and analysed. A program written in C reads the position of the sprayer from a GPS receiver and correlates it with a GIS file containing relevant spraying information. The GIS software used to create the files is usually ArcGIS software from ESRI(TM) but the program is also capable of reading an Idrisi Kilimanjaro file format from Clark's Lab(TM). The program finds and inspects all pixels under the spray width of a particular nozzle during one second of travel based on the previous direction and speed of travel of the sprayer. The sprayer nozzles are turned on and off using solenoids valves. The solenoid valves and the GPS system use a 12-volt power supply so the entire system can be operated on the electrical system of a vehicle such as a tractor. The GPS receiver is a John Deere Starfire and, with the SF2 differential signal, has an accuracy of plus or minus 10 cm. The pixel resolution varies from 1 meter to 18 centimetres.;The control program also creates an output file to be used by an analysis program. The analysis program displays the area of the field that has been sprayed along with the cost savings associated with the precision spraying.;Results show that, to achieve the desired spray deposition, a time factor needed to be added to the control program. The GPS positional error associated with tree lines typically found on blueberry fields does not significantly affect the capability of the system to precisely deliver agrochemicals. Achieving a substantial cost savings on pesticide application is possible but, based on the GIS maps received from the Applied Geomatics Research Group and the cost associated with their creation, the chemical cost savings do not warrant the use of the precision sprayer. Provided that the delivery of the GIS maps from aerial surveys could be automated, this system could be economically viable. On the other hand, further research and development could produce an on-the-go system that makes use of spectral sensors that would generate the same amount of savings and eliminate the need of a GIS map.