Menu
Menu
QUT roboticists develop world first robot to destroy starfish damaging Great Barrier Reef

QUT roboticists develop world first robot to destroy starfish damaging Great Barrier Reef

Robot to help automate searching and eradicating harmful crown-of-thorns starfish, as there are too few human divers to do the bulk of the task

Two Australian roboticists have developed a new underwater robot to tackle coral decline across the Great Barrier Reef caused by the crown-of-thorns starfish (COTS).

An estimated 40 per cent of coral decline across the UNESCO-listed World Heritage Site has been attributed to the starfish.

To help counter the impact of the species, Dr Matthew Dunbabin and Dr Feras Dayoub from the Queensland University of Technology developed COTSbot.

The robot comprises stereoscopic cameras for depth perception, five thrusters for stability, GPS and pitch-and-roll sensors. An arm of compressed air that injects the harmful COTS with bile salts is what makes the robot unique.

Computer vision and machine learning underpin the robot's ability to identify the species among coral.

The robot has been given thousands of still video images of COTS taken by divers to learn from. The system was trained and developed for six months.

"Its computer system is backed by some serious computational power so COTSbot can think for itself in the water," Dayoub said in a statement.

The robot is designed to help with the laborious task of searching for COTS and eradicating most of them across the entire 2300 kilometres of the Great Barrier Reef. The robot will be able do the bulk of the job, with human divers tackling the remainder.

"Human divers are doing an incredible job of eradicating this starfish from targeted sites but there just aren't enough divers to cover all the COTS hotspots across the Great Barrier Reef," Dunbabin said in a statement.

The robot can work underwater for eight hours at a time, equipped with up to 200 bile salt injections to destroy starfish.

"The COTSbot becomes a real force multiplier for the eradication process the more of them you deploy - imagine how much ground the [COTS eradication] programs could cover with a fleet of 10 or 100 COTSbots at their disposal, robots that can work day and night and in any weather condition,” Dunbabin said.

If the robot is unable to identify a starfish with a certain level of confidence, it takes a photo and sends it to a human expert to verify, with the robot then recording or putting into memory the class label (crown-of-thorns starfish or not) of the photo.

One of the challenges the roboticists faced when feeding the robot with images was the complexity of underwater environments and the varying visibility and colour changes depending on how deep under water a starfish is.

A breakthrough at James Cook University with the development of a one-shot injection method that is as effective as about 20 injections spurred the roboticists to develop their computer vision and machine learning project for detecting COTS.

Dunbabin, who began work on the system 10 years ago, had been stuck on how to effectively get a robot to inject a crown-of-thorns starfish.

"That was the game changer that opened the doors for a robotic solution to the COTS problem. Combining this with new advances in machine learning meant we could make COTSbot a reality."

The robot has undergone its first trial at Moreton Bay for testing its navigation and parts, with the next trial taking place later this month. The roboticists will test it on real crown-of-thorns starfish. Initially a cautious approach will be employed, with a human verifying each COTS the robot identifies before it injects them with bile salts.

The roboticists are planning to deploy COTSbot to work across the Great Barrier Reef in December, but need funding to scale up manufacturing.

Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.

Join the CIO newsletter!

Error: Please check your email address.

Tags robotQueensland University of Technology (QUT)Great Barrier Reefcomputer visionmachine learning

More about James Cook UniversityQueensland University of TechnologyTechnology

Show Comments
Computerworld
ARN
Techworld
CMO