So, with crocodile season ramping up right across the top end, it’s good to see Sydney’s University of Technology (UTS) catch croc fever with its CrocSpotter Ai.
What’s it do I hear readers ask? Is it an artificially intelligent crocodile that cruises Australia’s north protecting swimmers from dangerous crocs? No, it’s basically an app that spots crocs.
Now while spotting crocs is a solid and popular pastime for tourists heading up to the Northern Territory and Far North Queensland, what UTS have done is develop an algorithm with Amazon Web Services which allows a drone to livestream croc footage over a long distance.
It may not sound like much but CrocSpotter Ai can help conservation organisations and with emergency services to track crocs in real time. What it means is rogue or curious crocs can be captured once spotted and moved where human contact is less likely.
How does it work? Ultra-low latency streaming is relayed via the app over large distances to a monitoring team far away from the actual crocodiles.
In tests completed by UTS, teams at two locations achieved a successful live video stream from a known crocodile habitat on the Mowbray River in North Queensland. It worked by fitting cameras to a drone which beamed the video stream over 1600 kms back to base.
The drone carries CrocSpotter Ai as part of its payload with an onboard video. As video streams in real-time from the drone to the pilot on the ground, the algorithm operates by ‘washing’ the video, and alerts the pilot to a possible threat below.
The threat is highlighted immediately by a flashing red box around the detected animal, drawing the pilot’s attention directly to that part of the screen. For those safely back in the monitoring room, this red flashing hotspot can immediately warn authorities of an imminent crocodile threat in a particular region. The potential to use the AI right across the top end is enormous especially when you consider the algorithm driving the AI has a 93 per cent accuracy of identification.
As for the other seven per cent, it’s up to the individual being warned to make up their mind if it’s a croc or not. It could be a log for example – a potentially nasty log with a muscular bite.
Professor Michael Blumenstein and Dr Nabin Sharma from the UTS Faculty of engineering and IT who first worked with Westpac Little Ripper to develop SharkSpotter, the Ai-powered drone-based technology for protecting our beaches and keeping our swimmers and our wildlife safe from sharks, have formulated the croc algorithm.
The Queensland Government who much prefer the slogan “Beautiful one day, perfect the next” as a tagline as opposed to “beautiful one day, chewable the next” are keen to see CrocSpotter technology rolled out for crocodile season from Mission Beach to Port Douglas.
Reason being is crocs like all water. They love the beach, they love river estuaries, mangroves, freshwater streams, remote creeks and even ponds and if you’re a holidaymaker in Queensland during the summer months, they’re the exact places you want to hang out yourself – maybe not the mangroves.
What you don’t want to find or meet while swimming is an angry saltwater crocodile although most crocodile hotspots are clearly marked with warnings about not swimming or even fishing.
Professor Blumenstein said that the speed of the cloud-based AI can spot crocodiles in real-time, which is a world-first and a technological breakthrough given the very low latency.
Dr Sharma said the technology enables crocodile detection in complex environments, including murky and muddy waters in both wetlands and the open ocean.
“This is the first time this sort of animal detection drone technology has been deployed via a high-quality video stream,” he said.