Australia has introduced WeedScan, a pioneering artificial intelligence (AI) weed identification, notification and management application, a "game changer" in the country's approach to tackling weed-related challenges.
This innovative mobile and web-based application enables users, including farmers, land managers, local councils, Landcare groups and general citizens, to easily identify, report and map priority weeds by simply uploading a photo of the suspected weed.
WeedScan is a collaborative effort, supported by the Centre for Invasive Species Solutions, the national science agency CSIRO, the New South Wales Department of Primary Industries (DPI) and the governments of South Australia, Queensland and Victoria.
The app is trained on more than 120,000 weed images - making it a quick task to spot and manage those pesky weeds.
Introducing WeedScan! ????WeedScan is Australia’s first AI weed identification and management app. Developed with @CentreInvasives, @nswdpi, @sagovau, @QldGov and @VicGovAu, it's for identifying, reporting and mapping priority weeds in Australia.
???? https://t.co/lp5zNz4K6Z pic.twitter.com/Ja4YJwJBWW
— CSIRO (@CSIRO) December 12, 2023
Economically and environmentally beneficial
Andreas Glanznig, CEO of the Centre for Invasive Species Solutions, emphasised the app's potential, stating: "Weed management is notoriously difficult due to the challenge of correctly identifying weeds.
"WeedScan addresses this by enabling quick identification of weeds, many of which are new and emerging, and not widely known."
Highlighting the economic and environmental impacts of weeds, which cost producers $4.9 billion annually, Glanznig underscored WeedScan's role in combating habitat reduction for native species, environmental degradation, invasive species proliferation, waterway obstruction and potential bushfire risks.
"Game changer"
Dr Hanwen Wu, principal Weeds Research scientist from NSW DPI, called WeedScan a potential game changer.
"This app is unique in its ability to automatically identify weeds, provide instant local weed management information, and notify local weed control authorities," Dr Wu remarked.
Dr Alexander Schmidt-Lebuhn, senior research scientist at CSIRO, elaborated on the AI model's robustness.
"Our world-class AI model, trained on images captured across Australia and tested by various users, including farmers and community groups, can identify hundreds of weeds. Users just need to upload a photo for a quick, AI-assisted identification," explained Dr Schmidt-Lebuhn.
WeedScan is hailed as a significant tool in protecting Australia's natural environments and agricultural landscapes, offering a free, user-friendly digital solution for efficient and effective weed management.