by Alicia Gutierrez Brown

The Australian Forest and Wood Innovations (AFWI) Centre for Sustainable Futures, housed at the University of the Sunshine Coast (UniSC), has awarded $1.5 million in funding to support a new research project that will accelerate automation in Australia’s plantation forestry sector. 

The project, titled SilvaNaut: Incorporating Autonomous Operation into Australian Forest Machinery – Robotic Weed Control Conditions, directly supports AFWI’s theme of making the most of our available wood fibre by enabling safer, more efficient and higher-yield forest operations.

SilvaNaut will develop and field-test ‘The Autonomous Forest Navigator,’ a bolt-on autonomous navigation module that retrofits existing forestry machinery to operate hands-free between plantation rows. Designed initially for young (0-3 years old), open-canopy plantations, the system will replace manual joystick control with precise autonomous navigation capable of operating in rough, real-world forest terrain.

Project Lead Daryl Killin said the core challenge the project addresses is safe, reliable navigation in environments that are far more complex than agricultural settings. The project will be delivered through Daryl’s company, Native Conifers Carbon Sink, who registered the first tree planting project for carbon credits in Australia under the Carbon Farming Initiative in December 2012. The company has been exploring remote-controlled weed control for a decade. 

“There’s a real need for an autonomous vehicle that can do weed control in place of humans or human-driven machinery... but unlike driverless cars or broad-acre agriculture, we’re talking about remote locations, uneven ground, harvest residue, and a very high need for precision, because if you spray the wrong chemical in the wrong place, you can kill the tree,” Daryl said. 

The technology integrates high-resolution Global Navigation Satellite Systems (GNSS), LiDAR, inertial sensors (IMU), and AI-based obstacle detection, building on proven control technology developed by project partners. This geospatial backbone will enable not only robotic weed control, but also future applications such as inventory assessment, fire management, and forest monitoring.

Australia’s forestry sector is under growing pressure from labour shortages, rising operational costs, safety risks, and increasingly stringent environmental expectations. Daryl said manual weed control is labour-intensive and often undertaken in hazardous conditions, particularly in the early years of plantation establishment.

“Not many people want to put on a knapsack and work in remote areas with snakes, spiders and rough terrain anymore," he says, "And even when people are available, you’re limited by human constraints; you can’t work at night, reliability varies, and safety risks are always present.”

The Autonomous Forest Navigator removes operators from high-risk terrain while improving consistency and efficiency. Importantly, it enables round-the-clock operation, allowing growers to target optimal spraying windows that are often missed under current manual systems.

“Weeds in the first two years are critical,” Daryl said.

“If you lose 200 trees out of 1,000 per hectare early on, you’ve lost future options for wood volume, and you can’t put those trees back later. That loss shows up 25 or 30 years down the track, right when the return on investment really matters.”

While automation is well advanced in agriculture, Daryl said many existing solutions are over-engineered, too costly, or poorly suited to forestry conditions.

“Agricultural systems are often designed for flat land and annual crops,” he said. “Forestry is a much longer game. We needed a forestry-specific solution that’s cost-effective, robust, and fit for purpose, not something adapted from agriculture that doesn’t quite work.”

The SilvaNaut project aims to fills this gap by providing a vehicle-agnostic, bolt-on autonomous system, compliant with international technical standards and designed for mid-sized forestry equipment. The system is also being developed to work alongside drones, enabling smart task allocation between aerial and ground-based weed control.

“There’s still an important role for ground-based rigs... our system is designed to work with drones, not compete with them, choosing the right tool for the right job,” Daryl said. 

The project is being delivered through co-design with seven major Australian forestry companies, ensuring strong industry relevance and a clear pathway to adoption. 

Project partners include:

• HQ Plantations
• HVP Plantations
• Australian Bluegum Plantations
• Australian Carbon Farming
• Forestry Corporation NSW
• Forest Products Corporation
• Midway Limited

In addition to the commercial forestry partners, Native Conifers Carbon Sink – with support from the Regional University Industry Collaboration Program – engaged James Cook University to develop a complementary aspect of the technology’s development, which is training the SilvaNaut to identify young plantation trees from weed species, and to be interoperable with current industry drone-spraying technology. The project also has a Trans-Tasman element, supported by New Zealand-based companies Lincoln Agritech and Wrybill Robotics, who have significant existing technology that will be fast-tracked in Australia, as well as the New Zealand Forest Research sector. 

Field trials will commence in Queensland, with additional trials planned across Australia. Performance will be evaluated against navigation accuracy, labour savings, safety outcomes, fuel use, and overall cost-benefit compared with manual operations.

Professor Mark Brown, Director of the AFWI Centre for Sustainable Futures and UniSC’s Forest Research Institute, said this project meets a key AFWI goal of making the most of Australia’s available wood fibre by supporting our national capacity to meet growing demand for sustainable timber and wood products. 

“By improving early-stage plantation management, this project will help narrow the 30-40 percent gap between biological potential and realised wood yield, bringing together industry experience, AI expertise and real-world testing, to build something foresters can actually use," Mark said.

The project also supports workforce renewal by creating new roles for operators in forest robotics, remote operations, and data-driven management, helping attract a more technically skilled workforce to the sector.

“This is about giving forest growers confidence to adopt automation in a way that makes sense for forestry,” Daryl said. “If we get weed control right early, we protect future yield, improve safety, and make better use of the wood fibre we already have.”