Automation - the way of the future
Coined in the 1950s, ‘Artificial Intelligence’ (AI) is no longer the domain of science fiction. It’s well and truly here and fast becoming mainstream.
In fact, if you’ve used Google Maps, Spotify or Siri, you’ve used some form of AI, perhaps without even realising.
In a recent report by the AI Forum, AI is defined as ‘ Advanced digital technologies that enable machines to reproduce or surpass abilities that would require intelligence if humans were to perform them’.
Titled Artificial Intelligence: Shaping a Future for New Zealand, the report estimates that by 2035, AI has the potential to increase New Zealand’s GDP by up to $54 billion.
It lists the top five drivers of AI uptake as the technology’s ability to make sense of vast amounts of data, automate tedious or dangerous work, support decision making with speed and accuracy, reduce business costs by automation, and optimise business processes.
A white paper released in March by Callaghan Innovation predicts that AI will have ‘‘an extreme impact’’ on New Zealand’s agriculture sector in the next decade.
‘‘In agriculture AI is driving automated harvesting and milking, smart monitoring of plants and stock health, as well as more precise forecasting of crop and dairy yields,’’ says Callaghan Innovation chief executive Vic Crone.
‘‘Artificial intelligence is not just a technology for tech companies,’’ she adds.
‘‘In New Zealand and around the world, businesses across diverse industries are using it in a host of ways to create serious competitive advantages.’’
In the white paper, Thinking Ahead: Innovation Through Artificial Intelligence, the authors write:
‘‘We believe there will be an extreme impact in agriculture, with many tasks able to be automated and optimised. Presently we see the bulk of this impact occurring in the next two to five years … focused on efficiency and cost savings.’’
One of the ways in which AI is being used is in the application of machine learning (a term often used interchangeably with AI) to near infra-red images of soil samples by researchers at the University of Waikato.
This means the samples don’t need to be sent to the lab – so faster turnaround times – and it also enables farmers to apply fertilisers ‘‘much more efficiently’’, according to the report.
Callaghan Innovation also foresees robots roaming between plots, collecting eliminating weeds.
‘‘These bots are light enough that they do not damage the soil, and because they release herbicides only onto the weeds, they’re also more environmentally sustainable.’’
The paper also outlines the ways AI can be used to improve efficiencies in livestock management through the optimsation of feed and the dispensing of medicine. The technology can constantly monitor stock movements, eating patterns and health, and immediately flag animals that are showing unusual behaviour or reduced wellbeing.
Another example given in the paper is the prediction of crop yields well ahead of scheduled harvests.
Even stabilising a device as simple as a smartphone to ride along on trailers can help provide better estimates, suggest the authors.
‘‘The device videos the orchard on the go and can then geo stitch a map.
‘‘From there, video-based machine learning systems can detect, and therefore count, seeds and fruit months in advance.
‘‘Having these insights in advance would also enable harvesters to undertake sectionbased optimisations, improve food safety, and direct fertiliser to specific locations.’’
Professor Ian Yule, director of the Massey Agritech Partnership Research Centre, believes we are just at the beginning of seeing AI making a real contribution to the agricultural sector in New Zealand.
Along the lines of the C-Dax pasture robot, which was developed to help measure pasture autonomously (and showcased at the National Fieldays last year), Professor Yule predicts that in five years we will see basic robots on farms doing simple tasks such as recognising and spraying weeds, and picking and packing fruit. Much of the AI technology is now coming out of the research phase and into the development phase, he says, citing the purchase of Blue River Technology late last year by John Deer. data and