Real-time variable rate Pix® application system using a plant height sensor

Abstract

The objective of this study was to develop a chemical application system that could measure plant size, determine the optimum chemical rate to apply and control that application. A plant height sensor, the MEPRT growth relationship software, and the Raven chemical injection system were successfully integrated into a real-time variable rate chemical application system. The height sensor was 45% accurate in predicting plant heights within +/-3.8 cm two weeks after the pinhead square growth stage, but the heights were systematically overestimated by an average of 10.4 cm at the first white flower growth stage. It could not be determined from the data where the offset originated, but the height trends down the rows were generally tracked. With the Raven application system, the actual carrier flow rate lagged the desired rates. It was later determined that instead of using a 1-second update rate, a 3-second update rate would allow the control system to reach the setpoint before a new desired rate was received. A cotton growth regulator (Pix) was applied at uniform and biomass-based variable rates in a replicated study that included two cotton fields. Plant mapping data was collected during the season. Plant heights were significantly shorter in areas that received Pix than areas that did not. The only parameter that showed a statistical difference at the α=.05 level was the number of open bolls in Field B, 83 days after emergence. The lack of water for the growing season likely contributed to the lack of significant differences in final yields harvested between the different treatments. Since the yield differences between the treatment and control sections were mostly negative, this translates into financial losses from Pix application for the 2000 growing season. Other advantages can be envisioned from the use of the real-time variable rate system for other chemical applications. The sensor could be used in insecticide applications of fast growing areas of the field, or for nitrogen application based on deficiencies in the plants. This study is only the beginning for numerous possibilities associated with real-time application systems.