Predictive Analytics

Crop Type Classification

We create global crop and land use masks at various granularities — field to national — identifying field boundaries and classifying the planted area for key crops such as wheat, barely, soy, canola and corn.

Field boundary detection

Field boundaries are a prerequisite for field-level assessments. Applying a deep learning approach to the raw RGB and NIR Sentinel-2 data, we have been able to automate this process.

Yield Prediction

Provides an in-season prediction for expected yield at varying spatial resolutions, ranging from field-level to national level.

Supply Chain Vision

variable rate nitrogen

This map identifies areas that need more or less fertiliser in the crop at farm or regional level across a large landbank.


Compare fields to measure their performance either spatially or temporally.


Our Desiccation Zoning tool identifies areas of non-senesced vegetation and targets desiccation applications accordingly. This targets non-senesced crop, volunteers and weeds.

Variable Rate Fungicide

This service allows you to select the number of zones across the field and decide the chemical products rate per zone to precision target your chemical applications more accurately.

Green area index

Map the Green Area Index of your whole field and at farm and regional scale.

Canopy coverage

Mapping the percentage of potato canopy coverage per unit area, in order to optimise and monitor crop growth and aid management decisions.

Low cost, high frequency imagery at scale

Daily satellite imagery to help businesses monitor their upstream food supply chains.



Harmonisation and digitalisation of sustainable agriculture frameworks

land use

Classifying land use using satellite imagery to verify conservation and habitat projects


Identify crop rotations as a component of sustainable farm management practices.


This tool infers crop residue from the soil reflectance at different wavelengths in the spectrum, and classifies tillage type accordingly.

cover cropping

Using a threshold of NDVI as an indicator of bare vs covered soil, we can detect the use of cover crops using satellite imagery.


Provides an estimation of renewable organic matter.

Precision Agronomy

Plant counting

Quantifying vegetation by measuring the difference between near-infrared (reflected by healthy vegetation) and red light (absorbed by healthy vegetation).

sizing grading

Automatically draws the contours of each plant in the field to calculate their size accurately.


We create a prescription map identifying the areas where weeds are present. Spray only the weeds in the paddock - and not bare ground.

Contact Us

We are aware that someone has been impersonating Hummingbird Technologies with offers of employment. We are taking this matter very seriously and our lawyers are taking the necessary steps to end this. Meanwhile, genuine emails from the company will only come from an email address.

Website by Big Orange Media © 2021 Hummingbird Technologies Limited