As Bill Gates observed (in 1996), “We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten.”
Let’s take the opportunity to reflect on what ‘tech’ has changed in the mango industry in the last 10 years and speculate on coming changes.
A technology that has made a huge impact to mango crop management in the last decade has been in irrigation control.
Irrigation automation has been made possible by radio telemetry and smart sensing precision technology, for example, soil moisture probes and bore level monitoring devices, resulting in increased efficiency through reduction of inputs like fertilisers, fuel and labour.
This has allowed smarter and more cost-effective farming, especially important with price pressures and labour shortages The technology also helps to protect the environment as fertiliser applications are not lost through the soil profile through over-watering.
The labour invested in manual tap turnover can be redirected to focus on maintaining irrigation infrastructure and ensuring each tree gets watered. It also assists in watering during the night and on weekends when staff may be unavailable.
Bore level monitoring devices ensure bores are protected by measuring drawdowns at regular intervals and triggers an alarm if there is an issue. This helps irrigation scheduling, protects the asset and assists in the sustainable use of the water source.
Soil moisture probes are now much more reliable than they were 10 years ago and the software for visual translation of where the applied water ends up and recommended watering volumes is now much more user friendly. There are also several providers, which is a sign of maturation of the technology.
Handheld near infrared spectroscopy (NIRS)
It has been interesting to watch this technology evolve.
NIRS enables non-destructive estimation of dry matter content of mango fruit. Dry matter content is an index of starch and sugar content in the fruit, which in turn is related to the eating quality of the ripened fruit, as all starch converts to sugar. The measurement has proven to be useful in judging both harvest maturity and eating quality.
Australian developed technology became available on pack-lines two decades ago, but it was not adopted in the Australian mango industry. Sorting after harvest effectively created an inferior class of fruit of reduced value.
OneHarvest championed the use of handheld NIRS as it became available, with measurement of Calypso fruit on-tree. This information is used in the decision-to-pick, with a delay in harvest allowing an increase in fruit dry matter content (increasing at about 0.1% w/w per day, depending on growing conditions).
Today the technology has spread through the mango value chain – large farms have their own units, small farms are served by AMIA industry development officers, and marketing groups and retailers have units.
Retailer product specifications have been amended to specify dry matter specifications, varying by cultivar. As for any technology, handheld NIRS continues to improve, with a recent major advance being the introduction of Artificial Neural Network models in the operation of the units, which are proving to be much more robust in use than the prior method.
Development from flower to harvest-ready fruit is a function of time and temperature. ‘Heat units’ were introduced to the Australian mango industry by Yan Diczbalis in the 1990’s so the general concept is not new.
However, temperature can vary between blocks across a farm, depending on geography, so use of a temperature record from a far-away BoM station or a single station on the farm limited the potential of the technique.
Field sensors with data loggers became available but the job of downloading and transferring data became an extra task for the farm manager. In recent years, low field sensors with LoRa (long range radio) communication have become available. A web app (at www.fruitmaps.info) provides automated estimation of harvest date from user-entered flowering dates from ten-year history data and current season temperature data of a local LoRa enabled sensor.
The SensorHost sensors, made in Rockhampton, feature a battery life of over two years and a range of several kilometres.
Machine vision flower and fruit load estimation
Estimation of the timing of a flowering event is fundamental to estimation of harvest timing using heat units. The level of flowering is also useful in the first estimation of potential fruit load for the season. The actual fruit load will reflect pollination conditions and level of fruit drop, with a reasonably reliable estimate possible at stone hardening stage, approximately 5+ weeks before harvest.
Measurements of flowering and fruit load can be done manually, but this requires a consistent operator.
Machine vision can be used to count mango panicles and to count fruit load from stone hardening onwards. One such system from the Multiscale research project is being developed as a product by Freelance Robotics. These systems have potential to include machine vision measurements of size of fruit on tree to forecast fruit size distribution at harvest.
A part of the ‘mango madness’ of the harvest season is in the management of a large number of casual harvest workers for the short harvest period. As the years passed, local labour disappeared, and the industry has become reliant on overseas labour.
At first this was based on backpackers, with nationalities drifting over the years, and then the Timorese and Pacific Islander labour schemes, accelerated by the COVID border closures. Mechanisation is needed to reduce harvest labour, with the same technologies that allowed mechanisation of the packhouse over the last decades relevant to harvest.
Mango fruit spurt an acidic sap when the stalk is detached. In the rest of the world mango fruit is harvested with stalk on, then transported to the packhouse for destalking.
The Australian mango industry has moved to snapping the fruit from the panicle at harvest, using harvest aids in the orchard that wash the fruit to remove the acidic sap, and convey fruit into field bins. Initial versions of the harvest aids were tractor towed, but now most units are self-propelled.
The next step in the process of mechanising harvest is to add mechanical arms to automate picking of fruit from the tree to the harvest aid. This is a technology to watch in the next years.
Technology innovation and adoption has helped to shape and strengthen the mango industry… and the future is looking even more “Mangolicious.”
*Written by Martina Matzner an independent technology adoption advisor, who submitted a thesis on mango production as part of her university studies in Germany before coming to Australia, where she became involved in the development, and later management, of a large mango enterprise outside Darwin, over a period of 25 years.