Blog Artificial Intelligence delivers solutions to farmers

University students test the GoMicro app accuracy

University students test the accuracy of the GoMicro mobile application

GoMicro

A collaboration between the Alliance and tech start-up GoMicro demonstrates practical applications of phone-based AI technology. 

By Gianpiero Menza 

It is an exciting time, as the innovation ecosystem and investors’ interests are largely focused on sustainability, discovering deep science solutions and creating companies to address challenges related to the environment, agriculture, and nutrition. 

The newly launched Alliance Accelerate for Impact Platform is systematically scouting and exploring collaborations with agri- and food tech startups to establish mutually beneficial partnerships and to advance our research and deliver greater impact. By giving a voice to founders of exciting and promising solutions, we have created a space for scientists to gain more exposure to the innovation ecosystem.  

Harnessing phone-based AI  

During the Science Seminar held on Friday, 3 September, CGIAR scientists had the opportunity to learn from GoMicro CEO and Founder, Sivam Krish,  who developed a high accuracy phone-based AI app used to assess grain quality. Building high accuracy AI apps is usually both expensive and time-consuming, as it requires not only AI expertise but a large number of images for training. The phone-based imaging technology developed by GoMicro has changed this. It is able to capture high-quality well-illuminated images of pests, plant diseases and mineral deficiency with a simple mobile phone attachment which makes it possible to build high accuracy (~80%) apps with just 50 images. Even greater accuracy can be achieved with more images.  

  

Further applications  

This technology can make it possible for researchers to build and test their own diagnostic apps. The GoMicro platform can also deliver the apps to a wider user base if required. This technology has been used to successfully identify the fall armyworm in Kenya by working collaboratively with five scientists based in Kenya, India, Sri Lanka and the UK. With this technology, early detection of the fall army worm is now possible and it is possible to identify the various stages of development with high accuracy.  

  

GoMicro is collaborating with CGIAR and the Alliance of Bioversity International and CIAT to assess grain quality using the same technology and is interested in collaborating with scientists from all over the world to build and deploy phone-based detection apps utilizing phone-based attachments that will eventually cost USD $1.  

  

If you have a promising solution that could benefit from CGIAR and Alliance’s collaboration, please drop us an email on[email protected]  

  

Visit the Accelerate for Impact Platform and get in touch:

Gianpiero Menza, Senior Partnerships & Innovative Finance Officer: [email protected]