Smart electrostatic crop spraying using remote sensing technology

Abstract

For this thesis, smart spraying robot was designed, constructed and tested to validate the concept of smart pest control. Electrostatic charging of sprayed pesticide was realized in a spray nozzle design that improved plant coverage and reduced wasted pesticide as well as soil pollution. A thorough investigation into electrostatic spraying was conducted, which was accompanied by extensive simulations and experimentation. The results obtained from the simulation experimentation on industry standard electrostatic spray system (ESS) nozzles along with laboratory testing of these nozzles, detecting spray coverage using water sensitive paper and additional optical spray visualization methods gave the necessary insight and experience required to develop a new spray nozzle. Additional COMSOL simulation and experimentation were carried out on a Fan Hydraulic Spray Nozzle (FHSN), the results of which allowed for the effective addition of electrostatic induction capabilities, thereby transforming the (FHSN) into Electrostatic Induction Spray Nozzle (EISN) which is one of the prime parts of the smart spraying system. SOLIDWORKS software was used in the designing parts of this nozzle which were then manufactured using a 3D printer. An AL05D robotic manipulator and a TTRK tracked platform from Lynxmotion ™ were the mini mobile robot components selected for the feasibility study of the smart electrostatic crop spraying system. This mobile robot was equipped with a CCD digital camera, a range detector, and path mark detector to provide the necessary sensors required by the smart electrostatic spray system. A Windows™ based mobile computer in addition to an ARDUINO™ based orksmicrocontroller system were chosen to provide the computational power required by the system. These were arranged in a master – slave configuration, with the main processing for images and motion being conducted inside the master computer using programs created by Matlab smart ™ software. The execution of motion commands and the operation of the range and path mark detection along with operating the spray nozzle were performed on the slave computer using C as the programming language. The manufactured smart electrostatic spray system moves along cotton crop rows with a camera that scans the selected plant for pest infestation on the upper and lower surfaces of plant leaves. When a pest is detected, the spray nozzle is targeted on it at the appropriate distance, and a burst of pesticide destroys it. The results of experiments have shown that using the electrostatic induction system improves coverage 3 to 4 fold and reduces soil contamination by 2 to 4 fold. The system has plenty of room for performance improvement, and future development will make it adaptable for application to other crops and applications.