Select Page

Michael Gomez Selvaraj, Crop Physiologist at CIAT

Behind the farmer mobile apps, workshops and shiny interface of EcoProMIS, a creative and dedicated team are working to collect and process data. The team are developing our data platform, which is unique in the breadth of data that is collected, including crop information, greenhouse gas emissions, farmer interviews, and satellite and drone imagery.

Birds Eye View

Across our pilot sites in Colombia, our colleagues are recording all of this information. One of the most exciting parts of the job is the flying of drones above farmers’ fields to capture high-resolution images.

At EcoProMIS, our fleet of Unmanned Aerial Vehicles (UAVs, the technical term for drone) make regular flights to capture these images. They are not standard photos, but have a high spatial and high temporal resolution, captured by special cameras and sensors attached to each UAV.

Our fleet of drones includes octocopters and quadcopters, vehicles with eight and four rotating blades respectively. By using drones, we can collect data in a non-invasive way and with greater accuracy and cost-effectiveness than historical ‘boots on the ground’ data collection.

This is cutting edge technology and together with the other data inputs gives EcoProMIS unrivalled understanding of each farm.

 

CIAT team with one of the Unmanned Aerial Vehicles used for field observations

 

Image Analysis

The UAV images collected in each field are sent to the CIAT phenomics platform, a computer that can process and ‘stitch together’ all of the images. As the science partner on the EcoProMIS project, CIAT are based in Cali, Colombia, where their team of scientists merge the data from thousands of high-resolution images.

To merge and analyse these images, they use fully automated software. This is CIAT’s Pheno-i image analysis framework and provides the solution for processing the high volume of raw images.

From the result of the analysis, the team are able to extract vegetation indices, a form of information that can indicate plant health and productivity. Correlations can be made between vegetation indices and key crop agronomic traits, which provides the information required to build the crops models and then pass on the knowledge to support farmers.

The drone images are further enhanced by combining them with satellite images of the farms. This is another exciting part of our work, to be covered in more detail in a future blog article, and is delivered by the UK company Pixalytics. As a project funded by the UK Space Agency, the use of satellite data was a key component of the project design and appeal.

The result of this UAV and satellite data is a highly advanced and accurate product to serve the agricultural sector.

Knowledge for Farmers

The use of the UAVs allows us to gather sophisticated data which can then provide support to farmers in the form of a knowledge-rich mobile application.

The EcoProMIS team is currently developing the first of these ‘knowledge services’ for growers of rice and oil palm. These services, built on the drone images and other data, will predict yield and provide decision support to the growers.

Yield prediction is one of the most valuable pieces of knowledge for a farmer. With this knowledge at their fingertips, the grower can understand if the crop is performing well and if this is not the case to investigate and address limiting growth factors. Furthermore, as growers continue to interact with EcoProMIS, using the knowledge and uploading their own data, the crop model will improve in accuracy.

It is our intention that the drone data, combined with the other sources of farmer data, will provide a strong ally to growers for the shared ambitions of achieving food security and environmental sustainability.