Gordon the robot may have started life as a pile of scraps on the floor, but thanks to the work of scientist Guy Coleman it now has a promising career demonstrating how machine learning could be used for weed detection on farms in the not-so-distant future.
Like his robot, Mr Coleman had been headed towards a different life, in his case a career in medicine, until a year sabbatical as a weed research assistant convinced him to follow his passion for agriculture.
"It's important to have a job you enjoy, a career. And agriculture is my passion," he said.
"I completed a Bachelor of Science but majored in both anatomy and agricultural science as a way of appeasing my persistent passion for the industry, ignorant that agriculture could constitute a career path.
"I decided to take a year off and work at home and overseas as a research assistant in weed science, gaining more experience in the industry. Clearly, I never went back to medicine."
Now a precision agriculture and weed scientist at the University of Sydney IA Watson International Grains Research Centre in Narrabri NSW, Mr Coleman has been selected as one of two Australian delegates for the prestigious global Bayer 2019 Youth Ag Summit.
"The Youth Ag Summit really encapsulates everything I love about agriculture, the ability to have beneficial impacts on so many, the involvement of technology and importantly the global perspective," Mr Coleman said.
"Weeds are a ubiquitous global problem, and the Bayer Youth Ag Summit will allow me to work with ideas and people from around the world, and see how we can all benefit from collaboration and sharing of these ideas.
"I can't wait to build networks with other passionate agriculturalists, tackling agriculture's big issues."
Mr Coleman said his work at the grain research centre focused on investigating artificial intelligence, robots and lasers for use in plant based systems.
"I'm trying to develop and take some of these advancements and put them straight onto Australian farms to help manage weeds," he said.
"The key to it all is identifying weeds using machine learning, and individually targeting them, either with targeted chemical application, or potentially without a spray at all."
Mr Coleman said his project started by autonomously mapping plant breeding plots, using drones, remote sensing and GPS technology to map growth rates.
"Now I'm programming my robot Gordon to identify, map and track weeds within broadacre crops. The robot started last year as a few bits of aluminium on my loungeroom floor, so to see it used in demonstrations, for research and now weed detection, is very exciting," he said.
"Once the weed can be identified and the machine learning developed, the technology will be able to automatically spot spray or even target a laser directly on the growing point of the weed. This reduces chemical use, reduces costs and helps manage resistance.
"My work is also looking at alternate methods of weed control, such as targeted lasers and electrocution. While in the early stages, this has the potential to take herbicides out of the system all together and offer farmers alternatives when needed."
The 2019 Bayer Youth Ag Summit will take place from 4-6 November in Braslia, Brazil.
Start the day with all the big news in agriculture! Click here to sign up to receive our daily Farmonline newsletter.