Science Wednesday: crop monitoring across the grain belt

Roger Lawes and colleagues from the Graincast project have just published an overview paper in Crop and Pasture Science of how the Graincast technologies work together to deliver crop identification and yield prediction across the Australian grain belt.

Everyone in agritech can learn something by reading the whole thing, but I would like to point out some particular features:

• The author list reflects the Graincast project: it includes agronomists, data scientists, and specialists in remote sensing, devops and social science. (A particular shout-out to the two early-career authors, Franz Waldner & Yang Chen.)
  
  

• The Graincast app was an experiment in trading information that would be of value to farmers for ground-truth information that – once we built Graincast – would be valuable to other actors. There’s still a great deal to learn about this tactic.

• The project that the Graincast team started was not the project they delivered. By taking user engagement and human-centred design processes seriously – at a time (2018-19) when this was really unusual in an R&D agency – Roger and the team identified the need to pivot at least once, and executed that pivot.