Stage 2: Research & Development
By 2050 agricultural output must increase 70% to ensure inclusive global food insecurity with an ageing global population and no increases in the amount of farm land... Crop production must become more efficient.
Farmers lose on average 30% of each crop before it leaves the farm due to disease, pests and weeds. DHT has developed a better yield predictor and crop loss diagnosis engine which diagnoses the specific cause of crop loss; as in specifically: what pests, what weeds, what disease.
How does the solution work:
Step 1: A farmer flies a drone over their field. Any drone, Dark Horse Technologies software is hardware-agnostic, making it cheap, versatile and able to be offered widely. (We offer auto-pilot as part of the solution helping to ensure the farmers safety, quality control, and removes the need for UAV flight certification for the farmer to use the drone).
Step 2: After the mission, the farmer uploads the captured images to Canopy.
Step 3: The images are stitched together to provide a map of the field, they are analysed to predict yield and examined for signs of diseases, pests and weeds.
Step 4: Should any signs of diseases, pests and weeds be found the solution will diagnose the specific cause based on the image output.
Step 5: A report with Dark Horse Technologies findings is emailed to the farmer along with suggested solutions.
We continue to improve the technology to the point we can operate in remote locations with minimal internet or electric infrastructure. Likely going to means 'Intelligent Drones as a service', where a drone will fly from a bay station where the drones are maintained, to the farm, a survey is conducted before the drone returns to the bay station. The data would then be uploaded to Dark Horse Technologies solution as usual and a report emailed (or texted) to the farmer.
Dark Horse Technologies solution is unique because:
1 - Yield Predictor: Dark Horse Technologies' can count and size the individual kernels of wheat from drone gathered images. This takes the guess work and assumptions out of yield prediction and instead directly measures expected yield.
All other crop yield predictors only measure yield indirectly by measuring the relative health or "green-ness" of a crop, then applying a broad assumption of how correlated leaf health or "green-ness" is to crop yield. There are several issues with this approach, one being the rather large margin of error; and the other being the potential for other plants (e.g. weeds which are very green in colour) being included in the calculation. Additional challenges with the "correlated colour prediction" method are that it is less helpful for crops which change colour during the growing cycle (e.g. wheat).
2 - Disease Diagnosis Engine: Dark Horse Technologies' solution is able to diagnose the specific cause of crop loss, as in specifically: What pests? What weeds? What Diseases? are affecting a crop, from drone gathered images.
No other crop health analysis engine does true disease diagnosis. Every other platform that claims to be diagnosing crop loss events is either:
1: Correlating input data at scale to make a "best guess" as to what is affecting a particular crop, or
2: Involves taking physical samples of the affected crop, making it a non-scalable solution.