A new weather platform built specifically with local crop models and hyper-localised weather data will soon be in the hands of Australian farmers.
The technology from IBM uses data from The Weather Company with the aim of helping farmers become more productive, improve yield forecasting and increase profitability.
Head of The Weather Company for Australia, Jamie Azzopardi, said The Watson Decision Platform was an industry-specific solution based off a platform called Operations Dashboard.
"[It] has a similar look and feel to a weather app that you would be used to using on your phone," he said.
Pairing the weather app technology with IBM's PAIRS Geoscope data processing engine, Mr Azzopardi said they had created an agriculture-focused platform with the capability of producing a "digital twin" of paddocks.
"This is an engine that can ingest huge quantities of data from satellites, IoT, weather information, drone footage, you name it, and in real-time processing based on the queries that are put to it," he said.
"Based on the crops they're growing - we have an algorithm that is built specifically for those crops localised to Australia - the producer can not only get very accurate weather forecast observations, but they can drill into their particular paddocks."
Mr Azzopardi said barley growers would be the first to use the AI models designed specifically with local factors including environmental, weather and plant biological conditions such as irrigation management and pest/disease risk.
By the end of 2019 this will expand to corn, wheat, soy, cotton, sorghum, cane and potato crops.
Mr Azzopardi said being able to deliver these solutions to producers who are trying to manage a whole raft of issues was only going to add value to them, but it had to be simple.
"The challenge is data in and of itself is good but it doesn't really add a lot of value, so what comes from the data is what do you do with it, how do you interrogate it, what insights and information can you get out of that data, what are the gaps that you need to complement with other data sources, and then how do you actually deliver all of that quickly, efficiently, easy to consume, and then is it cost effective."