When AI goes bananas: an app helps farmers grow healthy fruit

Bioversity International turns to clever algorithms for food security

by Max Smolaks 12 August 2019

A team of researchers from Bioversity International in Africa has created a smartphone app to help banana farmers protect their crops against diseases and pests.

The Tumaini App (meaning ‘hope’ in Swahili) is based on artificial intelligence algorithms that have been trained to recognize five major diseases and one common pest affecting the world’s favorite fruit, demonstrating accuracy of more than 90 per cent in most models.

The software has been tested in Colombia, the Democratic Republic of the Congo, India, Benin, China, and Uganda.

Tumaini can recommend the means of addressing a specific disease and automatically upload identification data into a global database to help coordinate international response. It is hoped that the app can stop disease outbreaks and protect the livelihood of small, independent farmers.

The paper describing the algorithms has been published in Plant Methods – an open access, peer-reviewed journal dedicated to the plant sciences.

“There is very little data on banana pests and diseases for low-income countries, but an AI tool such as this one offers an opportunity to improve crop surveillance, fast-track control and mitigation efforts, and help farmers to prevent production losses,” said Michael Selvaraj, crop physiologist at the International Center for Tropical Agriculture and lead author of the paper.

Bioversity International is a global research-for-development organization headquartered in Italy, working to ensure global food security since 1974. To create the app, the team – which also includes researchers from India’s Imayam Institute of Agriculture and Technology, and Texas A&M University – first collected 20,000 images of the ruined crops.

The app was then trained to recognize the nature of the disease using three different convolutional neural networks in combination with deep transfer learning.

Researchers say that, unlike the existing crop disease detection models which need clear, isolated pictures of specific parts of the plant, Tumaini can detect symptoms on any part of the crop and is trained to be capable of reading images of lower quality, inclusive of background noise.

“The overall high accuracy rates obtained while testing the beta version of the app show that Tumaini has what it takes to become a very useful early disease and pest detection tool,” said Guy Blomme from Bioversity International. “It has great potential for eventual integration into a fully automated mobile app that integrates drone and satellite imagery to help millions of banana farmers in low-income countries have just-in-time access to information on crop diseases.”