The OWL (Open Weed Locator) Project is an innovative, grower-led initiative driven by the Warren Cauliflower Improvement Group in Manjimup, Western Australia, with supportive seed funding through the VegNET innovation program and regional and global partnerships; The OWL Project – Manjimup is advancing affordable, open-source ag-tech solutions for precision weed management.

At the centre of the project is the OWL camera system, developed to detect and manage weeds in vegetable crops using AI-powered “green-on-green” and “green-on-brown” spot spraying algorithms. By addressing both environmental sustainability and cost accessibility, the project is designed to help growers reduce herbicide use, cut input costs, and make precision agriculture more accessible, affordable and achievable on-farm without the barriers of patents and high costs associated with commercialised products.

One of the key features of the OWL Project is its practical, hands-on approach to grower engagement and technology adoption. With leadership from inventor Guy Coleman, Manjimup growers Warren Cauliflower Improvement Group recently participated in an international industry tour, visiting the open-source technology hub at Copenhagen University in Denmark. This was followed by a series of local open-source focused workshops, farm and field demonstrations in the use of AI weed detection and spraying technology, including guided regional farm visits that utilise the technology in mixed production, onions, and spinach.

As part of the OWL Project’s international collaboration and knowledge-sharing efforts, a dedicated field day was hosted at the University of Copenhagen’s research facilities in Taastrup. The event brought together researchers, growers from Denmark and Australia, ag-tech innovators, and industry experts to explore the potential of open-source technologies in advancing precision agriculture globally and showcasing some of the technology available.

The Field Day began with a series of presentations highlighting the growing role of open-source tools in farming. Guy Coleman from the University of Copenhagen introduced the open-source movement in agriculture, followed by Patrick Coleman from Pemberton.digital, who explored the practical applications of open hardware in ag-tech. Contributions from Frederik Larsen of Agrovi and Australian grower Jake Ryan provided valuable agronomic and farmer perspectives on the adoption of tools like the OWL system in real-world vegetable production.

The program also featured insights into various autonomous cutting-edge precision technologies with speakers presenting on collaborative projects such as OneCropHealth (OCH), satellite-based predictive systems from Geco Engineering, drone solutions by CropUp, tractor-based innovations from PerPlant, and precision implements from EcoRobotix. A mapping demonstration by RoboWeedMaps further showcased the integration of data for weed management collecting data sets via drones.

In the afternoon, attendees participated in hands-on field demonstrations. Technologies including the OWL system, and a DIY spot-spraying setup with a 12m boom. Commercial sprayers were displayed alongside open-source and other higher cost solutions from PerPlant, EcoRobotix, CropUp, and RoboWeedMaps. The demonstrations provided a tangible look at how open-source and precision technologies can be applied directly on-farm, reinforcing the day’s theme of accessible, farmer-led innovation in weed control and crop management.

A full day OWL workshop was delivered by Guy Coleman, which guided growers to assemble their own OWL camera units using Raspberry Pi-based hardware and learn to train AI models on weed image datasets. These sessions provided foundational skills in building, coding, and deploying the technology for real-world weed detection and control, with growers commenting how easy it is to assemble with the guidance by the OWL Project and accessible information via GitHub and YouTube tutorials.

The OWL Project group in Manjimup, led by Jake Ryan, are pioneering a 16-channel platform initially targeting brassica crops, and planning farm field demonstrations and a podcast to come once the unit is fine-tuned and can be shared with other growers. OWL is laying the groundwork for broader application across regions and crop types.

The OWL Project’s next phase focuses on establishing an open-source, publicly accessible weed image library. With three core pillars—industry study tours, hands-on hardware workshops, and the weed image dataset—the OWL project has evolved into a scalable platform for grassroots innovation.

A full report on the industry tour will be featured in the next Australian Grower, with insights from the farms and growers who participated in the eight-day open-source project tour.

FIND OUT MORE
Please contact Katrina Hill on 0427 373 037 or email katrina.hill@vegetableswa.com.au