RapidAIM have developed a highly accurate Fruit Fly Forecast, equipping growers with more knowledge than ever before on pest pressure in their region
What is the RapidAIM Pest Forecast and how reliable is it?
The RapidAIM Fruit Fly Forecast is updated daily and can predict pest pressure in the region with 90% accuracy. Growers will be able to see the forecasted pest pressure over the next 5 days, along with the actual pest pressure for the past 14 days.
The forecasted pest pressure is relative to the season so growers can see how the current pest pressure compares against detections for previous months.
The whole aim is to show growers exactly what’s happening in the region in real time, to take the guesswork out of crop protection.
This graph shows our model’s predictions (red) versus the actual detections that occurred (blue).
I’ve seen other forecasts before – how is this one different?
RapidAIM has hundreds of sensors rolled out across several horticultural regions in Victoria and Queensland. These sensors have been collecting real time detection information for the past 12 months.
Our engineering team have then taken this information and collated it with granular weather information for the region and by using machine learning have created a highly accurate predictive model – a bit like the BOM.
As more and more detection information is received the model is refined to be increasingly accurate.
This is superior to other models, because of granularity of input information. We know the exact location and time of day of every fruit fly detection. Whereas other models are based on information gathered by manual trap counts which are conducted on a weekly or fortnightly basis. At best, using this information, will generate a basic trend over time and isn’t hugely reliable.
The RapidAIM pest forecast is extremely accurate. The graph above shows our model’s prediction versus the actual number of flies that were detected, so growers and agronomists can now plan pest management action rather than having to react when there’s already a problem.