The end may be nigh for invasive weeds if the work being conducted by scientists at James Cook University (JCU) involving the creation of a weed-killing robot proves successful.
Electrical Engineering post-graduate student Alex Olsen has been working on the project since April 2015 under the guidance of Professor Peter Ridd as part of his PhD in Engineering and Related Technologies at JCU.
Mr Olsen said the scale of the invasive weed problem is such that there is no chance of it being solved by manual weed spraying alone.
His aim is to develop a machine that is capable of differentiating between weeds and other plant life using an algorithm to detect a range of variables including colour, shape and texture.
“We try to think of it as what we do when we first see it with our own eyes,” he said.
“We can identify green straight away, then determine the shape of the plant, then the leaves, and then we strip that down to a point where we can grab small texture windows for a feature comparison to find out if it matches what we’re looking for.”
The machine will have a series of cameras on the front and herbicide sprayers on the back, which will work together through a series of processes to determine precisely when to spray the weed.
The machine was trialled at Hidden Valley on the western slope of the Paluma Range near Townsville, and the team was encouraged by the early performance of their prototype.
He said the team envisaged the device being towed by, or fitted to, existing agricultural vehicles in the short-term, but the possibility existed that it could be fully autonomous in the future.
“Due to the implications of the rough terrain it’ll be used in, we decided the robot would be more effective if it was operated manually for the time being, we are also currently tweaking the speed so it can run at at least 10km/h so its task can be carried out more efficiently.
Mr Olsen said weeds are the starting point for the technology but it also has scope for use in agriculture or plant classification.
“It’s not just for weed removal, it could be for helping to identify which species of plant are located in a national park.
“The algorithms are for image processing, if you have a knowledge of this it can apply to anything.”
In the meantime Mr Olsen said the team is refining the image recognition technology so that the robot is spraying onto weeds with at least 90 per cent accuracy.
“We’re also currently working out how we can get to operate via solar energy instead of being powered by a car battery as it currently is.”
He said he’s hoping that the robot will be made commercially available by the time he finishes his PhD degree in 2017.
