Orcharding and the power of (more) data
The report cites disruption due to global trends such as population, climate change, biotechnology and smart technologies and industry-specific change accelerators including consumer preferences, emerging technologies and changing value chain and firm configurations.
As a consequence of the identified disruption drivers, the report says the agricultural ecosystem faces new challenges, from which three significant growth opportunities for current and future players can be derived:
1. Improving yield efficiency
2. Increasing supply chain efficiency
3. Decreasing complexity along farmers’ value chain. Much of the analysis focuses on feeding the growing population, and the big industries of wheat, rice, corn and soybeans. What does the disruption mean for orcharding?
I suggest the same trends and opportunities apply. Perhaps in fruit and wine production we are less interesting in feeding the world, but we certainly want to increase productivity, cut losses, increase value capture and do it all with less complexity.
Many new tech factors will contribute to improvements. For
now, let’s have a quick look at the information revolution.
Joseph Aamidor of Greentech Media compared industries and the potential benefits of the Internet of Things IoT revolution. He notes that compared to building management and manufacturing, many farms are starting at a very “data-light” position. This may change rapidly, with farm data platform OnFarm suggesting the average farm will generate 4.1 million data points per day by 2050, up from
190,000 in 2014.
How many of you are generating (much less storing) 190,000 data points per day in 2017? How are you going to collect and manage your 4 million data points? Where are they going to come from?
First of all, we expect the resolution of data capture to increase, both in time and space. So we’ll start collecting information about every plant every time we go through (or over) the orchard.
Imagine a set of sensors on the tractor (yes, I think you’ll be using them for a while) that capture canopy, fruit and soil information as you mow or spray (and record what sprays went where at what rates). Climate information too – the temperature, humidity, water status and light around each plant.
Combine that with satellite and UAV maps, connected sensors delivering climate, soil and plant data every few minutes, labour records in plant by plant detail and multiply that by blocks and seasons and years. New agriculture jobs will have titles like sensing technician, GIS officer and chief data scientist.
What is some of the information that would be most useful to you now? As you develop more in depth understanding, what information will you want to collect later? How will you do that?
Think about yield prediction. That information is vital for the grower to balance crop loads and estimate yield, and for the exporters and sectors for planning logistics and marketing. How good is current practice?
Our predictions are based on a relatively small amount of data captured by people trying to accurately count buds and shoots and flowers and fruits. We know from many trials just how variable the results from this type of collection are. But it is the best we have, so we extrapolate this to guess whole orchard and whole packhouse and whole sector production. And we are often far from the mark.
Emma Leonard of SPAA reported (“Precision Viticulture in the Riverland”) that wine makers want an automated system that throughout the season predicts final yield to at least 95% accuracy. This compares to current systems that are at best 85% accurate at berry counting and 80% at bud counting. Local observations in the viticulture, kiwifruit and apple sectors suggest even poorer estimates are common.
If a machine could count buds, flowers and fruits, and measure the size of fruits as the season progresses, would these estimates be better? Given current accuracy, the machine hardly needs to be perfect! But is it feasible?
A number of groups in New Zealand and around the world are trying to develop technologies to count buds, flowers and fruits in apple and kiwifruit orchards and vineyards. Notably, all those I have seen are under development. Counting orchard stuff with a machine is very, very difficult.