AN ASSESSMENT OF SUGARCANE YIELD MONITORING CONCEPTS AND TECHNIQUES FROM COMMERCIAL YIELD MONITORING SYSTEMS

A NUMBER OF attempts have been made to monitor sugarcane yield variation within a block in Australia. These have ranged from the early yield monitoring systems based on discrete mass measurement, to the current focus of predicting yield via surrogate measurements based on chopper pressure, feed train roller displacement and elevator power. Recent work aimed at assessing commercially available sensors (Jensen et al., 2010) suggested that there were several areas in which there was room for marked improvement. Rather than testing commercially available sensors, this paper details the testing that was conducted on evaluating the measurement concept. These concepts included; the pressure drop across the elevator and chopper motors, a load cell in the elevator floor and the angle of opening of the top feed roller. These concepts cover those being employed in the commercial units, both past and present. Trials were conducted during the 2010 season in the Bundaberg region and in both the Bundaberg and Herbert regions in 2011. Campbell Scientific CR3000 dataloggers were used to read each of the sensors at 40 Hz and record the averaged value every second, along with the GPS information. In addition to this sensor data, sugarcane yield was also measured directly to determine the accuracy and resolution of the respective yield monitoring concepts. Yield was determined using two methods and included mill (bin) weight data for individually consigned sub-blocks and weighed 50 m row samples into a weigh truck. The approach used was consistent with the methodology previously developed to assess the accuracy of commercial yield monitoring equipment during 2008/09. Preliminary analysis indicates that there are considerable similarities between the yield monitoring concepts in terms of their ability to measure yield, and that how the sensor data are recorded and managed is critical to the accuracy and overall performance of these concepts as yield monitors. This paper reports on the findings of this work and makes recommendations for further refining the devices and for additional work.