Australian Feed Monitor Explained

• Get weekly estimates of pasture biomass and ground cover

  
  

  • What are my ground cover levels?
  • What are my paddock pasture biomass levels (kg DM/ha)?
  • How does my feed on offer (FOO) compare to last month?
  • How many grazing days do I have available in each paddock and across my enterprise?
  • When will the mob hit the target weight?

• What is the Cibo Labs PastureKey Service?

  • The Cibo Labs PastureKey service combines world-leading satellite remote sensing and data science with on-farm knowledge to estimate feed supply for every hectare and to monitor changes in the feedbase and land condition on a weekly basis.

• How Does The PastureKey Service Work?

  • A “living laboratory” model of agricultural data science tailored specifically to the Australian landscape and livestock industry.
  • The service uses Sentinel-2 satellite imagery captured every 5-days around the earth by the European Space Agency (ESA).
  • Thousands of GPS-located observations of total standing dry matter (TSDM) collected by Cibo Labs and producers using a mobile App are analysed using machine learning algorithms to deliver within-paddock estimates of TSDM across the entire farm.
  • The service integrates world-leading data science into some open and easy to use web mapping, data and mobile Apps to support day-to-day grazing management decisions.

  .

• Save time on more objective pasture assessments

  • Weekly imagery allows more targeted and objective pasture assessments, saving significant time.
  • Paddocks can easily be prioritised to optimise grazing rotations.
  • Detailed pasture yield assessments are undertaken using dedicated GPS-based Mobile Apps which are provided to clients to support calibration and validation efforts. 
  •  

• A Living Laboratory continually refining predictions based on on-ground knowledge

  • The GPS-based field data collected by Cibo Labs and producers is used the train prediction models based on 5-day satellite data using Machine Learning and high performance computing to “tune” the models for local conditions.
  • We work with producers to coordinate data collection/calibration and to improve predictions

   

• How accurate are the predictions?

  • We currently have over 7500 GPS-located field sites being used to train the model.
  • Currently predicting over a range of ~8000kg/ha the median prediction error is <200kg/ha with an r-square correlation of 0.85. (needs updating)
  • Importantly, landscapes and pasture systems are highly complex and there will often be anomalies where more field data is needed to improve predictions (there are no silver bullets). In these situations we work with producers to coordinate data collection to improve predictions.

• A new estimate for every paddock and the entire farm every 5 days.

  • Summary statistics on total standing dry matter (TSDM) in kg/ha and Feed on Offer (FOO) in total kg/paddock are generated for every paddock and provided for every satellite capture.
  • Data quality flags are also provided when the data is affected by clouds.
  • The TSDM and FOO predictions are pushed automatically into third party systems including AgriWebb and PairTree. Simple CSV and XLS forage budgeting spreadsheets are also provided.

• Why Is It Important?

  • Maximise profitability and sustainability through managing stocking rates correctly
  • Save time and improve decision-making through objective information on the state and condition of every paddock on a weekly basis.
  • Get the information you need to support day-to-day grazing management decisions, as well as seasonal and long-term planning.
  • Provide regulators, supply chains and consumers with objective and transparent data to demonstrate your sustainability credentials.