Theofrida Maginga, a PhD candidate at the African Center of Excellence in Internet of Things (ACEIoT) at the University of Rwanda (UR) specializing in Embedded Computing Systems is in advanced stages of developing a ChatGPT-powered Swahili chatbot that will help smallholder farmers detect crop diseases quickly and easily.

Dubbed ‘Mkulima GPT’, it will integrate Artificial Intelligence (AI) with Internet of Things (IoT) technologies to support farmers with useful agricultural information in a culturally-sensitive manner.
It is funded by Bill and Melinda Gates Foundation and was developed in collaboration with Dr. Jimmy Nsenga, an affiliated honorary lecturer at the ACEIoT.

With this technology, smallholder farmers with limited literacy and scarce resources will be able to quickly and easily detect the crop diseases such as Northern Leaf Bright. The team will also integrate artificial intelligence (AI) and Internet of Things (IoT) and use no-invasive sensors to monitor non-visual early disease indicators including volatile organic compounds (VOCs), ultrasound movement and the uptake of soil nutrients.
Speaking to The New Times, Maginga compared the platform to “a virtual extension officer.”

“We are bringing a virtual extension officer with the assistance of the Chat GPT model, which is a very famous application right now. The farmer can get advice in real time and can be able to communicate with the technology in the local language,” she said.
The project is currently in its testing phase, where the developers are encouraging potential consumers to try it out and provide feedback on what can be improved.

It already has a website and a dedicated WhatsApp phone number to which farmers can send their queries either by text or audio, in Swahili or English, to get information regarding how they can prepare their farms, maize disease management, post-harvest procedures, and so on.

After getting the funding, both researchers are now collecting as many questions and answers as possible that originate from farmers themselves which will be stored in a database and will be reproduced in terms of questions and answers as if farmers were communicating with extension officers.
doc516

“We have the IoT part, but instead of forcing farmers to report to extension officers when they detect diseases, the technology will do that minimizing interaction with human beings”, Maginga said.

In addition to having the hope to extend the platform so that it will cover more crops besides maize, Maginga emphasises the importance of applying ChatGPT and other AI-based solutions in local languages across various sectors
She reckons that Mkulima GPT will be expanded to more African languages.

“For a farmer to interact well with the technology, it has to be in their own local language. They find it more suitable and adaptive,’ she noted.

When the final testing processes get done, the platform is expected to be launched in December, but meanwhile, the developers are inviting more experts in agriculture and artificial intelligence to have a look at their work and share insights related to what can be made better, so that when it reaches the small-holder farmers, it will be ready for use.

Event photos