Drone Based Pasture Measurment
How It Works:
Ambient Lighting
The most critical factor in measuring pasture cover with drones is ambient lighting. To ensure accurate data and reliable future predictions, light reflectance must be measured using a light calibration panel. This allows the AI model to select the most appropriate reference dataset for accurate analysis.
Ground-truth data is collected from test plots measured with a rising plate meter. The AI model then learns the relationship between the light reflectance in the imagery and the corresponding plate meter measurements, forming the basis for its predictions.
Its All About The Data
Initially, test plots are measured in every paddock to capture the range of pasture types and conditions across your farm. The AI then learns the unique characteristics of each paddock, allowing it to accurately predict pasture cover with minimal ongoing input.
A master AI model further improves accuracy by comparing your paddocks with similar records from a wider dataset. Unlike satellite methods that rely on past growth and environmental data, especially under cloud.
Drone measurements base predictions on realtime visual observation. Over time, additional inputs like soil moisture and temperature can be added to further refine results.
The Outputs
Initially, pasture cover estimates will be generated using the master AI model to get you up and running quickly. Over time, as you gradually input data from your own paddocks, the system will develop custom sub-models for each paddock, learning their individual characteristics and behaviour.
Regular, automated drone flights provide continuous updates on pasture coverage, keeping the model current and responsive to real-world changes. The more data you contribute, the more accurate and reliable the predictions become, enabling better-informed management decisions.
At the same time, ongoing test plots and data collection keep you actively involved in the process, ensuring you stay connected with how your pastures are performing and how the model is evolving.
Extra capabilities:
Break Fencing
Set up grazing breaks that allocate equal amounts of feed for your animals. Simply input the number of animals, their daily intake requirement, and how many breaks you want. The system will then calculate and provide the exact locations where break fences should be placed to ensure precise and fair feed allocation.
Residual alerts
Continuous, autonomous drone flights can monitor a mob throughout the day. By setting a target residual, you can ensure animals are moved off a paddock at the right time — preventing overgrazing and enabling more precise pasture management.
Usable Feed
By analysing target and residual figures, you can understand what pasture is truly available to your animals. Before-and-after drone flights reveal how much feed was actually utilised versus wasted, helping the system build more accurate predictions of pasture utilisation for future grazing events.
Fertiliser
Over time, the AI model monitors paddock growth and identifies underperforming areas, enabling targeted fertiliser application where it is needed most. This reduces wastage and improves overall pasture utilisation. Variable-rate application maps can then be uploaded to advanced spreaders, which automatically adjust fertiliser rates in real time.
Weeds
Paddocks can be mapped in higher detail to pinpoint areas that require targeted weed control. The AI model identifies weed-affected zones and generates prescription maps for sprayers, ensuring chemical is applied only where it is needed.
Irrigation
Create water prescription maps for irrigators that tailor application rates to the specific needs of different areas within each paddock, ensuring water is applied where it is needed most.