Higher quality mechanized agricultural operations can be achieved with the use of the Global Navigation Satellite System (GNSS) signal positioning tools (correction signals), allowing a higher accuracy, which is extremely important to reduce operating costs and waste of inputs, in addition to allowing a more effective pest control. Thus, the aim of this study was to evaluate the behavior of the mean execution error of the positioning and pass-to-pass design in the operation of sugarcane ratoon cutting and insecticide application. Furthermore, the efficiency of controlling Sphenophorus levis through non-automatic steering (NS) and use of autopilot (RTX and RTK correction signals) in a sugarcane production plot of an experimental area located in the city of Motuca, SP, Brazil were evaluated for a total of 150 points by means of the statistical process control, analysis of variance, and descriptive statistics. Fipronil was the insecticide used for S. levis control. The evaluations consisted of the measurement of the mean execution error of the project during tractor operation in ten strides and five replications, in addition to the pass-to-pass (parallelism error) error between strides of the tractor-ratoon cutter assembly. In all strides, the mean execution error and mean error of the tractor-seeder assembly were within both the acceptable limit and the stipulated by the signal manufacturer, with values lower than 3.8 cm. The control charts were efficient to evaluate the behavior of RTX signal quality, facilitating the visualization within the limits of the project execution errors and pass-to-pass, in addition to contributing with an S. levis control 27.16% higher than the conventional control in the cutting operation of sugarcane ratoon.