Navigation control of a mobile robotic system in semi-structured and dynamic environment for controlled dose delivery of pesticides

Abstract

A major portion of the autonomous mobile robotics research has been in the development of reliable means for navigation and control of autonomous systems in a complex, irregular, unstructured and dynamic environment for farm duties such as seeding, pruning, pesticide spraying, reaping etc. Several practices have been established by numerous researchers in the motion planning of mobile robots. This paper presents an ultrasonic multi-sensor fusion based technique for navigation and control of the autonomous mobile robotic system for collision avoidance, target detection and canopy mapping. The autonomous system is a combination of real time monitoring and navigation, detection of the target and canopy mapping which delivers controlled dose of pesticide in a targeted manner according to the information gathered from the sensory attributes of precision farming data set through online and real time monitoring of the objects. Real time monitoring and navigation control, complex algorithm, data acquisition and storage, calculation and processing are designed with motors and sensors interfaced with microcontroller. Test results showed that the ultrasonic sensor fusion technique was capable of performing satisfactory guidance in dynamic environment and desired paths. The sensor fusion navigation was capable of compensating large error and resulting in satisfactory agricultural vehicle guidance.