Next generation weed management (BY23002)
Title: Next generation weed management (BY23002)
Project start date: 25 March 2025
Project completion date: 30 April 2030
Key research provider: La Trobe University, Nufarm Australia & Applied Horticulture Research
Summary: This program is tackling one of horticulture’s most costly and persistent problems, weed management, through innovative, technology-driven research. Weeds can slash crop yields by up to 50% and are particularly problematic in crops like potato, onion, garlic and leafy greens. With herbicide resistance increasing and chemical control options declining, the program is focused on developing and accelerating adoption of new, sustainable methods of weed control that are both effective and commercially viable.
The program covers three individual projects:
Applied Horticulture Research is partnering with grower-led groups in Queensland and Victoria to accelerate the adoption of AI-enabled robotics and autonomous vehicles for weeding, planting and spraying. The project focuses on real-world integration of these emerging technologies into Australian vegetable farms, providing support services, production system adjustments and clear economic analysis to reduce the risk for wider industry uptake. Through on-farm demonstrations, economic modelling, and industry consultation, the project aims to show how robotics can cut costs, boost productivity and reduce environmental impact, while helping growers future-proof their operations.
La Trobe University is leading a research initiative to investigate Anaerobic Soil Disinfestation (ASD), a chemical-free method of weed and pathogen control gaining traction overseas. By manipulating soil microbiology using organic materials and creating low-oxygen conditions, ASD has the potential to suppress weeds while improving soil health. The project will test this method across multiple climates and crops, including strawberries, tomatoes and capsicums, with trials in Victoria and the Northern Territory. The outcomes will provide Australian-first data on ASD’s agronomic, environmental and economic performance, supporting its use as a viable alternative to traditional fumigants and herbicides.
Nufarm Australia, in collaboration with Croplands and Kilter, is developing an autonomous robotic precision spot-spraying system that uses advanced machine learning and green-on-green weed detection to apply microdroplet herbicides directly onto individual weeds. This smart spraying technology is designed to reduce herbicide use by up to 95% while increasing yield and lowering labour costs. The project will deliver a comprehensive dataset, new weed detection models, and a fully automated robotic platform tailored for Australian horticulture.
Together, these three projects are set to deliver practical, forward-looking weed management solutions that can help Australian growers reduce chemical inputs, manage herbicide resistance, and stay competitive in a changing farming landscape.