Here’s how to tackle the menace of crown-of-thorns starfish – a robot, state of the art animal recognition engine, and 200 lethal shots of bile salts!
Queensland University of Technology researchers have developed COTSbot – short for Crown-of-Thorns Starfish robot – that is capable of spotting, tracking and hunting down crown-of-thorns starfish which are said to be responsible for an estimated 40 per cent of the Great Barrier Reef’s total decline in coral cover.
Specifically designed to hunt down crown-of-thorns starfish and control their population, the COTSbot can dive up to 8 hours at a time and deliver 200 lethal shots of bile salts. Its developers have fitted COTSbot with stereoscopic cameras for depth perception; five thrusters for stability; GPS and pitch-and-roll sensors; and a unique pneumatic injection arm to deliver a fatal dose to the starfish.
According to its developers, a COTSbot fleet will serve as great first responders and with the ability to recognise the starfish, this autonomous robot will be a great help in saving the corals of the Great Barrier Reef.
The robot’s advanced computer vision helps it to navigate in marine water conditions and its machine learning algorithms help it identify the starfish. In case the robot isn’t able to ascertain whether the animal it saw was a COTS, it captures images of the animals for visual inspection by a human later on. This is what makes the robot much more novel that it relies on human inputs as well when it is confused.
To ensure that COTSbot doesn’t go about delivering fatal shots to any marine animal, its developers have trained it using thousands of images of COTS collected on the reef. Its developers believe that COTSbot is the first autonomous underwater vehicle to be equipped with an injection system. It’s also designed to operate exclusively within a metre of the seafloor, one of the most dynamic and challenging environments for any robot.
Dr Feras Dayoub, who designed the COTS-detecting software, said the robot would continue to learn from its experiences in the field. Its computer system is backed by some serious computational power so COTSbot can think for itself in the water,” said Dr Dayoub, from QUT’s Science and Engineering Faculty and Australian Centre for Robotic Vision.
COTSbot idea spawned a decade ago
Dr Dunbabin has already been working on the vision system that can COTS some ten years ago, but the scientists gave up on the idea owing to the limitations in the the eradication methods used at the time. COTS were required to be injected some 20 times before it can be eradicated, and this is something that was too much for a robot in uncontrolled conditions.
Last year James Cook University (JCU) scientists came up with a one-shot injection method that had proved just as effective. This, Dr Dunbabin says, was a game changer and it motivated him to start working on the COTSbot again. Machine learning, and his vision system handled rest of the bit and COTSbot is finally a reality.
COTSbot is set for a field trip to the Great Barrier Reef later this month where it will be involved in a trial with live targets. During that trial, researchers are planning to verify each COTS identification the robot makes before the robot is allowed to inject it.
If things work out as planned, the COTSbot will be working the reef autonomously in December.