What is the WAVE Pilot?
WAVE stands for WA Vegetation Extent. It's a pilot study by the Department of Water and Environmental Regulation, designed to investigate the potential of machine-learning to map our native vegetation. The study involves trialling a number of approaches, to see if we can improve the way we gather data on WA's native vegetation, across our vast state.
In simple terms...
...we're training machine-learning systems to recognise native plants from satellite imagery. The challenge is for these technologies to accurately identify native plants across a variety of landscapes.
Ultimately, the process is helping us decide whether AI technologies are ready to play a part in the important work of mapping our native vegetation - and to what degree.
There's no wrong answer
Regardless of the outcome, the WAVE Pilot project is important for two reasons; Firstly, it's already advancing our capabilities in data and emerging technologies, and secondly, it's part of our essential efforts to ensure WA's diverse vegetation is accurately accounted for in policy and decision-making.
Why is WAVE so important?
Every day, government planners, industry project managers, Traditional Owners, and conservation bodies make decisions about native vegetation.
The issue we face is that WA is vast. It's hard to get comprehensive data on the extent of native vegetation in any given area, at any given time. This means decisions sometimes need to be made without adequate data. WAVE is a step towards more definitive data, and better-informed decision-making, accessible to everyone.
Scale demands innovation
Our state's incredible 2.5 million square kilometre landmass makes manual vegetation mapping impractical, especially as our maps need to be updated regularly to capture changes.
But now, for the first time, this pilot is testing whether machine-learning has reached the level of sophistication required to build accurate, automated vegetation detection tools.
Potentially, the pilot will lead to a new, comprehensive source of data, for the many organisations that need to account for native vegetation in their decision-making.
The outcome of the pilot is by no means guaranteed, but what we do know, is that the process itself is already moving our understanding forward, and putting us closer to a better way of mapping.
How it works
We're testing AI-driven image and pattern recognition technology that can identify vegetation and changes in satellite imagery.
Although the pilot is currently confined to three test areas, the hope is that the technology can be scaled up to map native vegetation across millions of square kilometres, identifying clusters of native vegetation.
- Our landscapes are nearly as diverse as our plant life, so we're testing and developing multiple data sources and processing methods across different landscape conditions, to see what works and what doesn't.
- Machine learning technology is not here to replace human decision-making. It's a tool to help us recognise and classify patterns in vast amounts of information, much of which is not visible to the human eye. The vegetation analysis outcomes provided by the tools will always be subject to human scrutiny.
- And, once the data has been considered, the responsibility for making a decision, is firmly back in human hands.
What will WAVE mean for you?
Depending on the findings of the pilot, potential benefits for different groups could include:
Government & policy
- Better planning data: Timely evidence-based decision-making with comprehensive vegetation information
- State capability: Building WA's innovation and data infrastructure
- Risk reduction: More consistent, dependable data reduces uncertainty in policy decisions
Industry & development
- Project planning: Better data for development proposals and approvals
- Reduced uncertainty: More comprehensive vegetation information available upfront
- Streamlined processes: Standardised, publicly available data potentially reduces assessment timeframes
Traditional Owners & community
- Supporting country care: Technology to support traditional land management
- Collaborative approach: Combining traditional knowledge and on-ground experience with modern detection tools
Conservation & environment
- Protection through knowledge: Better data for more effective conservation decisions
- Change monitoring: Potential to track gains and losses over time
- Evidence-based action: Publicly available data that carries more weight for conservation efforts
Research & technical partners
- Innovation leadership: Advancing WA's research and technical capabilities
- Knowledge contribution: Contributing to global vegetation mapping science
- Technical excellence: Rigorous scientific approach to capability building
The latest...
Right now, we're in Phase 1 of the pilot, testing different methods with different partners, to assess technical feasibility. We should have the results of this phase by November 2025.
Phase 2 of the pilot will then involve moving the best feasible idea - or ideas - into a proof-of-concept process for approximately 6 months.
Next Steps
The results of this testing phase will help inform a business case. The business case will recommend whether to go ahead with fully automated implementation, a modified partially automated approach, or we may conclude that current technology is not ready to take on this important task.
Regardless of whether full automation proves feasible, the pilot investigation is already building valuable knowledge and capability for our state.
Stay informed
Sign up to stay informed and we'll email you with progress reports, as-and-when information is available.
Contact Information
For general enquiries, technical questions or stakeholder liaison, please contact WAVEpilot@dwer.wa.gov.au.
Print this page
