Challenge

Cranfield University, a highly regarded agricultural research institute in the UK, was awarded Innovate UK grant funding for a project supporting smallholder banana producers in Colombia.

Smallholder farmers account for approximately 10% of the total banana production in Colombia. These growers are vulnerable to a number of externalities, including fluctuating commodity prices, rising labour costs and increasingly, climate uncertainties. In Santa Marta, Colombia, rising temperatures in particular have led to water shortages.

With this project, the Cranfield University team set out to develop an irrigation model tailored for use on banana smallholdings; a tool not available anywhere in the world.

Beyond the technicalities of developing the model, the logistics of delivering irrigation support in real-time to the smallholder farmers also posed a great challenge. The key issue in this project was that the recommendations had to be generated and disseminated weekly, and farmers needed to provide input data on a weekly on the actual amount of irrigation applied. This prevented the model from drifting and becoming useless.

Interacting with every farmer twice a week was simply impossible under the existing extension agent network available. The agents currently had the capacity to visit each farm once every 1-2 months. So communicating with the farmers directly was key if this project was to succeed.

Solution

The resulting downscaled irrigation model was combined with Climate Edge’s scheduled SMS feature, allowing tailored irrigation alerts to be delivered to every farmer in real-time. This provided a direct and personalised route to deliver the agronomic messages to each farmer, enabling the farmers to adapt their irrigation practices based on the recommendations provided.

Climate Edge also utilised their survey tool to ask farmers how much they were actually able to irrigate their plot compared to the recommendation. This crucial input ensured that the model remained accurate over time.

Results

The model generated by the Cranfield team provided irrigation recommendations which were 40% more efficient than the existing methods being used by the smallholder farmers. Recommendations were provided to a selection of pilot farmers during the project. After delivering the recommendations throughout the dry season for 8 weeks, an NPS survey was delivered to farmers using the platform, with 50% of farmers responding that the digital irrigation recommendations were either very effective or effective compared to their previous methods.