In May we undertook a crop assessment on our cherry trees - essentially an estimate of how many cherries we would produce this coming season. We used a traditional method of identifying an 'average' tree, counting the fruit buds, and making assumptions of the number and weight of cherries that each flower bud would produce.
With our thinking caps on, we quickly saw an opportunity for a more accurate method of gathering this data. We compiled a random list of 100 of our oldest trees, which will be three years old this summer, and had staff physically count every fruit bud on those trees. This gave us a more accurate census of how many fruit buds we have.
From these numbers, there are industry-standard assumptions made on:
How many flowers each fruit bud will produce,
How many of those flowers will be fertile and pollinated, and
Number of cherries those fertilised flowers will produce (fruit-set).
We determined an accurate crop assessment for our trees this season will be 579.3 kg per hectare - a total of 2091 kg of delicious cherries.
However - and this is where things get interesting - we have engaged and invested in Bitwise Agronomy in Australia, who are developing software that can do this for us.
Next years census will be done using our autonomous tractor mounted with cameras that film the entire orchard. Their software analyses the video taken and produces a full dataset of the orchard.
With automation, we can collect data on
Blossom count;
Pollonation;
Fruitset;
Crop size (tonnes per variety / hectare);
Cherry sizes;
Ripeness and harvest estimation times;
Probably a dozen other data points we have not yet thought of!
There are some interesting things happening in this space - check these guys out: