FIVE years have passed since CSIRO released its Graincast smartphone app, designed to forecast grain yield over a single paddock or an entire farm.
What happened to the popular free app, the first in Australian to offer easy-access, whole-supply-chain and commercially relevant information?
The technology is now licensed to an external party, Digital Agricultural Services, where it has been integrated into its cropping hub. DAS’ Crop ID product is being used to generate land-use maps for rural Australia and cites 75 percent accuracy in identifying at least 17 crop species, and up to 90pc in some areas.
CSIRO remains an equity shareholder.
The technology has also been further developed by CSIRO and now exists as the free Wheatcast service. This offers fortnightly wheat-yield forecasts during the growing season, both for Australia as a whole and for individual states – again, an Australian first. It also incorporates maps of soil water, water-limited yield forecasts and their uncertainty.
Before Graincast, there had been no comprehensive national system for quantifying yield and geo-locating crop areas with accuracy at a locally relevant scale.
Powerful handheld modelling
The app was revolutionary because it put powerful crop modelling in the hands of growers and agronomists; it was widely used and rapidly adopted.
It allowed advisers, bulk handlers, marketers, commodity forecasters and insurers to forecast grain production at regional and national level without the hassle of identifying soil types, calibrating models, searching for satellite information or trying to interpret complex data.
Every night it would run tens of thousands of complex model simulations to provide more than 400 independent grain farmers with daily estimated soil moisture and yield potential for any location in Australia.
Across 60 million hectares, Graincast made it possible to understand what was grown on every cropped field, understand what the yield potential of that field was, and understand what the tonnage produced from that field was.
In 2019, users were asked if Graincast information could alter cropping decisions. They nominated two situations: when timing topdressing later in the season, and as reassurance to commit to their plans, avoiding the urge to overreact to temporary weather events.
Graincast provided reliable forecasting, recording 0.15t/ha deviation, and an unbiased estimate of national wheat yield.
Forecasts were most uncertain at the end of April, and more accurate by the end of September.
Funded as part of the CSIRO Digiscape Future Science Platform, it used information from satellites, climate forecasts and sensors to estimate historical crop yields, yield potential and crop species.
Graincast was marketed as the only app giving instantaneous soil, water and yield information at any time without the need for substantial user input.
It was also an early attempt to capture crowd-sourced data; the designers were able to discover what crop was growing where.
Compared to other yield-prediction platforms, Graincast’s “human-centred” app was considered easy to use, and simple to input information. It could be asked multiple questions that could vary user-by-user and for a variety of purposes.
“Growers told us that they wanted something that was mobile, quick and easy to use,” CSIRO scientist Roger Lawes said.
“They didn’t want to be told what to do, they just wanted the key information so they could decide what was best for them,” Dr Lawes said.
“The app was easy to use, and growers needed to input just three pieces of information: the paddock they want analysed; the crop grown last season, and the crop they plan to grow or are growing in the current season; that’s it.
“We combined a broad suite of satellite-based crop-mapping, crop-modelling and data-delivery techniques to create an integrated analytics system that delivered a viable crop-monitoring service.
“This design process demonstrated that delivery of data as a service was viable at the field scale, as the end-user had the autonomy.
“But we realised that a vertically integrated data supply chain was required to develop crop-monitoring technology further.
“Digital agriculture is constantly evolving, and CSIRO is at the forefront of that change.”
HAVE YOUR SAY