Significant drops in the cost of remote sensors, better cloud-based server capacity and powerful artificial intelligence tech open up opportunities for New Zealand farmers to better understand their farm’s response to climate change and life-changing weather events.
Stephen Ponsford, CEO of Spark’s AI and data analysis company Qrious, says recent cyclone events have highlighted the role digital mapping and real-time sensing, to name two tools, can offer farmers to better prepare for future events.
Qrious has been working with two farm environment companies, MyEnviro and Adroit, on sensing tech in the Mangaone River catchment of Hawke’s Bay, providing the platform for real-time monitoring of water quality, emissions, fertiliser use and pasture growth.
“And we have just had an agreement with Hills Laboratories to do ground truthing on the real-time probes to calibrate their accuracy,” Ponsford says.
“For example, the sensors could detect an elevation in E coli levels, prompting the lab to intervene and test to check, so the sensors can act almost like a real-time digital smoke alarm for an event.”
After Cyclone Gabrielle, Ponsford says, there is potential for better risk management around future events, with farms able to combine real-time sensing and satellite footage and use AI to evaluate the risk characteristics of that particular farm.
It takes the work that catchment authorities have done at a modelling level for their catchments to a far more granular level that farmers can manage and account for in making decisions for future events.
“This has really only arisen in the past five years as we have seen things like IoT [internet of things] sensors drop in price to only $1-$10 apiece, and cloud-based AI approaching zero cost.”
It is the predictive analysis, the look at what might happen on farm given certain conditions, that Ponsford says is a particularly exciting step that comes after real time analysis of sensor data.
“You can then start on simulation analysis and get to develop ‘digital twins’ of your farm, with AI simply the only way to process the level of data this can generate.”
As weather events that were once one in a 100 years become more the norm, the need to understand their effect in the future becomes more critical than simply reviewing them afterwards in the belief they are unlikely to come around again any time soon.
He is optimistic the age-old bugbear of farm tech – lack of interoperability across different platforms – will soon be a thing of the past.
“At Qrious we have done a lot of work across other industries, including banking and insurance, where data sharing has been a requirement. On farms of course you also have IoT data to incorporate in the mix.
“We are going to start to see more done on verifiable data, data required for legislative purposes and for things like bank lending, which may require evidence of sustainability targets, for example.”
He sees the use of that tech becoming more transparent, sitting in the background of daily operations, ticking the compliance demand box simply by being entered in real time as part of daily operations.
At an industry processing level Qrious has turned its analytical capacity to help Zespri develop a maturity assessment system that offers a digitised quality control system for the organisation.
Ponsford says thanks largely to the ultra-fast broadband project and the Rural Broadband Initiative, New Zealand is fortunate to now have a solid set of digital arteries running through it, even more so than across the Tasman.
As more giant server farms are built in NZ, that resilience and the capacity to build agritech and export it offers an income-generating lever that puts less demands on physical resources than simply producing more food.
“We have a GDP weighted with high-carbon, lower-paying jobs, while our tech sector has a skills shortage.” Qrious has started a data academy as a talent pathway, offering $1 million in scholarships for women, Māori and Pasifika to build more local talent in the data analytics industry.
Research has shown more diverse tech teams are 50% more likely to collaborate and commit successfully.
