Fall ArmywormMachine learning, a variation of artificial intelligence that includes the development of algorithms that independently learn new information, has innumerable applications. An example of this can be found in Africa, where the fall armyworm pest in Uganda has ravaged crop yields. Amid the destruction, a new machine learning-based app created by a Ugandan developer has the potential to stop the spread of the crop-destroying pest.

Agriculture in Uganda and the Fall Armyworm

Approximately 22% of Uganda’s GDP comes from agriculture, with most Ugandans working in the agricultural sector, often engaging in subsistence farming. With the nation’s economic performance relying on successful agricultural harvests and the population’s everyday food source coming from their own crop yields, any invasion of pests in Uganda can have serious consequences.

In 2016, Uganda experienced its first invasion of the fall armyworm pest, the larva of the armyworm moth. A native of the tropical regions of the western hemisphere, the fall armyworm pest eats through crops for nourishment before its transformation into a moth. By mid-2017, the fall armyworm had been detected throughout Uganda and was estimated to have caused $192 million USD in maize crop losses alone. In some regions, up to 75% of crop yields were lost.

Despite the severe threat posed by the fall armyworm pest in Uganda, local developers have created a machine learning-based tool to assist Ugandan farmers with detecting the presence of the fall armyworm in their crops and preventing its spread.

Machine Learning to Protect Crops

In the aftermath of the arrival of the fall armyworm pest, Nazirini Siraji, a Ugandan woman from the city of Mbale, began work on a modern solution to the age-old problem of pest invasions. After attending one of Google’s Codelabs events, Siraji used Google’s TensorFlow platform to develop her Farmers Companion App. TensorFlow is an open-source machine learning tool that enables developers like Siraji to create digital solutions powered by artificial intelligence.

The Farmers Companion App enables farmers to use mobile technology to identify this specific pest on their crops and their lifecycle stage. Using this information, the app notifies the user about the threat level faced by their crops and the extent to which the fall armyworm has the potential to spread. The app also recommends specific pesticide treatments that can be used based on the situation of the farmer’s crops.

According to Google, the app has already been deployed in the agricultural lands surrounding Mbale, where Siraji partners with local farmers in utilizing her Farmers Companion App.

Big Tech Meets Local Developer

The global expansion of the internet has been accompanied by a rise in local innovation aimed at solving local issues. In Africa, pest invasions have been responsible for countless crop shortages and famines, which exacerbates problems of instability and poverty. While invasions from pests like the fall armyworm will inevitably occur in the future, they will not happen again without opposition from new technology.

John Andrikos
Photo: Flickr