Innovative research into fruit quality at harvest within an orchard block will make management, traceability and pack-out performance more effective for the apple, pear and stone fruit industries, Mark O’Connell* writes.

Agriculture Victoria is conducting scientific research to field test the accuracy of sensors on a harvesting platform to measure fruit quality (sweetness, maturity, firmness, diameter, skin colour, blemish, sunburn, shape and internal disorders).

The system has GPS location, enabling the spatial variation in fruit quality to be examined and provide data-driven decisions in subsequent seasons for spatial management such as pruning, thinning, leaf blowing, variable rate spraying and zonal harvesting where feasible and financially justified.

The project builds on foundational technology developed by Rubens Technologies – a commercial handheld spectrometer to measure fruit quality (rubenstech.com). The sensor array used on the platform harvester includes a fluorescence spectrometer, a reflectance spectrometer and an optical camera.

The project will also develop and evaluate an automated fruit bin tracking system (e.g., RFID reader) on a platform harvester and associated downstream data capture and analytics. Here, each individually harvested fruit is linked in the orchard against RIFD-tagged field bins so that fruit quality is captured prior to short or long-term storage and then passing through a commercial fruit grader.

Work is being undertaken to integrate field data (fruit quality and location) and packhouse data (fruit grader analytics) into a traceability application programming interface (API) service.

This will allow different applications (e.g., orchard spray records, fertiliser program, beehive management) to communicate with each other and improve data access and functionality, as well as provide improved efficiency and insights into orchard production performance.

Field testing and initial calibration on apple, pear, nectarine and plum occurred during autumn 2023. Optical images captured on the platform harvester show high levels of accuracy and precision for fruit size (and shape) and colour estimates. Fruit spectra data were compared well against spectra obtained from a handheld spectrometer.

Future development includes real-time feedback on picker efficiency (e.g., colour and size picking) —this would require instantaneous conversion of spectra land optical image data to fruit quality parameters combined with software so the harvest platform operator can be informed and provide feedback to the picking crew.

There is also the potential to field grade fruit that is damaged and/or does not meet market specifications.

Further details on the innovative study into fruit quality at harvest within an orchard block to make management, traceability and pack-out performance more effective for apple, pear and stone fruit industries are available on a YouTube video: https://youtu.be/UIztueyy32A.

Agriculture Victoria has partnered with Federation University to offer a PhD scholarship in Information Technology and Horticulture Production Sciences to study on the go orchard fruit quality sensing on the platform harvester for pome and stone fruit crops.

For further information, contact Mark O’Connell, Tatura SmartFarm, Agriculture Victoria.

*Mark O'Connell is a researcher at Agriculture Victoria, Tatura Smart Farm and Honorary Senior Fellow Centre for Agricultural Innovation at the University of Melbourne. Contact: mark.oconell@agriculture.vic.gov.au