Comparison between bin tonnages and yield monitor predictions

THIS PAPER DETAILS the investigation undertaken to track harvested cane from the field to the mill and to compare this information with yield monitor data. The testing was conducted late in the season in the Childers region in 2016. The harvester evaluated was a 2015 model John Deere 3620 that had been equipped with both a prototype yield monitoring system (Jensen et al., 2013) and a JD yield monitor. Only cane that was cut into haulouts that were a straight tip into the rail wagons was tracked. No topping up of bins was permitted. The bin numbers were noted and tracked to the mill. The flagging function of the yield monitoring system was used to identify the start and stop of material cut that was manually aligned with the bin numbers. The yield monitor data were processed using the yield monitor protocol (Bramley and Jensen, 2013). The resultant yield file was used to predict the amount of cane cut between the inserted flags in JD file and these values compared with the bin tonnages recorded at the mill. A total of 85 bin pairs (the bin configuration on each haulout consisted of 2 ? 6 ton bins) were cut during this evaluation. The chopper sensor performed well (R2 = 0.77) but the elevator sensor provided the best relationship (R2 = 0.86). The JD yield monitor was not evaluated as the work was ?commercial-in-confidence?. To fully overcome consignment issues, spatial data need to be transferred from harvester to haulout to bin to eliminate errors.