Decision-makers faced with environmental issues now have an artificial intelligence tool to make the science more accessible than ever
Jim McLeod first used aerial images to see what was happening in the bush in 1995.
“You could probably make out a house on it, but nothing smaller,” said the Waikato Regional Council scientist.
Inaccessibility, computing power and lack of public interest all restricted the flow of environmental data.
Nowadays, it comes in a flood.
McLeod and his team use satellite and aeroplane photography to survey forests for kahikatea trees and the spread of kauri dieback.
He says the issue now isn’t just getting the information - it’s analysing it.
A government-funded artificial intelligence programme at Waikato University promises to help councils and decision-makers make sense of huge swathes of environmental data.
The MBIE-funded, $13 million platform TAIAO gathers and collates environmental data from around the country and makes it available to scientists.
The computer programme will be able to automatically check DoC predator cameras, map at-risk native trees with satellite imagery and detect algal blooms on waterways - all with minimal human involvement.
University of Waikato Environmental Research Institute director Professor Karin Bryan said the platform, which launched last month, would change things dramatically.
“Decision-makers need to understand what’s happening before things go off the rails,” she said. “Once a lake is green, it’s hard to turn it back.”
More readily available access to real time data would allow councils and the Government to make more informed decisions more quickly, said Bryan.
Waikato Regional Council is already involved, using the platform’s ability to scour aerial imagery to find kahikatea stands.
“You can train the computer to recognise different signatures,” said McLeod. “A person could do it but it would take much longer. Now we can pick up a potential change in the trees much more quickly.”
Jannat Maqbool is the associate director of the new AI Institute at the University of Waikato, which developed the AI.
She said sporadic environmental data in New Zealand and quickly changing weather patterns due to climate change meant we needed a new approach to how decision-makers gather data.
“We need to be doing it live,” she said. “TAIAO can change and adapt and stream what is happening in the moment.”
Bryan said this is where environmental science meets data science, and claimed the platform would allow environmental scientists to make community consultation a more integral part of their work.
“Automating much of the data work allows the scientists to do the human work,” she said. “Whether that’s considering the Treaty or working with the local community.”
Greater and more timely insight into the environment would empower scientists to make the right decisions. Maqbool said this is exactly what the platform would provide.
“The environmental scientists will be able to focus on the things they are good at,” she said. “In that sense it will give them a bigger voice.”
After years of killer robots in movies dominating the public’s view of AI, Maqbool hopes TAIAO will lend AI a more positive image by helping communities stay informed about New Zealand’s quickly-changing environment.
She said they hoped to develop a te ao Māori approach, and put time into considering who owns the data.
“Hopefully that will set a new tone to AI, which has sometimes been depicted negatively - with it taking everything over.”
Bryan said promoting the accessibility of scientific data was of paramount importance, so the platform is available free.
TAIAO will run over the next seven years and is a collaboration between the University of Waikato, the University of Auckland, the University of Canterbury, MetService and Beca.
