BY LEE HICKIN
Like many Australians I have fond memories from my childhood of throwing in a line and hoping for a bite. It’s important that future generations will be able to create similar memories.
Right now, the oceans are facing profound environmental challenges as well as the stress of feeding a growing global population.
More than 60 million people are employed globally in fisheries and aquaculture and the industry is worth $US1.5 trillion a year. Around 17 per cent of the animal protein[1] consumed by people around the world is seafood.
As the population continues to grow – from 7 billion today to 9 billion by 2050 – global fish stocks will face even greater pressure[2], and ocean health is already threatened by pollution and climate change. When a dead whale washed up on an Indonesian beach recently it had more than 1,000 pieces of plastic in its stomach.
We have to act. We can.
Artificial intelligence (AI) and cloud computing are combining to offer powerful platforms that can help scientists tackle a huge array of challenges that the planet faces.
AI is increasingly being used to help monitor the health of our oceans and global fish stocks and it’s been a privilege for Microsoft to be involved in a ground-breaking program in the Northern Territory that uses AI to monitor the waters off Darwin and help scientists working in this critically important area.
What’s perhaps not widely recognised is that there are more coral species in the waters off Darwin than there are on the Great Barrier Reef, and species of fish and turtles that are only found in Northern Australia. A team of scientists working with the Northern Territory’s Department of Primary Industry and Resources have been exploring how to best monitor and manage marine health in this important and very beautiful part of the world.
The unique nature of the Top End means they’re not going to jump into Darwin Harbour in a wetsuit because of the crocodiles and sharks. Instead scientists have fitted underwater cameras onto buoys to capture video of fish in protected reefs.
In the past a scientist has then had to sit in front of a screen to scour the video to identify and count fish that swim past. Now using Microsoft Azure and the same sort of AI that allows facial recognition on social media, the video reviews are being automated – and thanks to machine learning, achieving 95-99 per cent accuracy.
The scientists can then take that data and act on it, freed from hours of relatively unrewarding work spent watching underwater videos.
Advances in the technology means that the first iteration of the solution was ready in a month and the whole system was up and running in just six months.
Microsoft is committed to our mission to empower every person and every organisation on the planet to achieve more, which is why we’ve developed this application using open source code and made it available on GitHub. It is now freely available to scientists around the world who might be able to deploy similar systems and, using machine learning, train them to identify fish specific to their part of the world.
Over time a much deeper understanding of global marine health will arise.
And the technology has broader potential. The Northern Territory is already considering how it could use similar set-ups to monitor feral fish in freshwater; to monitor the by-catch in commercial fishing vessels; and track cattle movements across the Territory.
It’s not just the Top End benefitting from AI.
Oyster farmers in Tasmania are using AI technology to monitor water quality and track conditions when oysters might become unsafe for people to eat.
A solution developed using Microsoft Azure and machine learning technologies by Melbourne based company, The Yield, is providing much finer and accurate control for oyster farmers – reducing their costs and reining in consumer risk.
The application of cloud AI to environmental and planet-scaled challenges is limited only by the imagination.
Microsoft has created the AI for Earth grant to promote success in achieving a sustainable future by funding the work of start-ups, universities, and other organisations advancing sustainability across the globe.
Why? Because we want future generations to be able to know the joy that comes from a tug on the line.
We owe it to ourselves; we owe it to the planet.
[1] http://www.fao.org/3/i9540en/I9540EN.pdf
[2] http://www.worldbank.org/en/topic/environment/brief/oceans
