Fighting animal disease and pest control with artificial intelligence and machine learning

Agripredict is a start-up company that provides artificial intelligence and machine learning as risk management tools for the agricultural sector. Its systems enable farmers to detect plant and animal disease, which, in conjunction with pest infestations can reduce farm yields.

Before its launch in 2016, the biggest technical challenge facing the company was the high cost of some of the training models available on the marketplace in conjunction with high data costs in Zambia.

“We opted for AWS because even then it was affordable, scalable and its accessibility was second to none,” enthuses Mwiza Simbeye, machine learning engineer at Agripredict. “As a start-up, we also received the value of one-year credits from AWS. This set the course for an astounding first year.”

In the field, a farmer takes a photograph of a plant or animal suspected of being affected by a disease. Once Agripredict receives the image its system will automatically give a diagnosis; options for treatment if required, and the location of the agro-dealer nearest to that farmer.

“We can achieve this because of AWS’s Sagemaker, EC2, Lambda, RDS, Amplify products. We also recently started implementing Elastic Beanstalk as well,” Simbeye elaborates.  “Our website is easily deployed via Amplify and we train our machine learning models by utilising Sage Maker, which then puts the models into production.”

Accessible

“Having increased efficiency means we can deliver our services to more people while keeping our costs low and accessible,” he points out.  “By using AWS’s EC2 our services are also available on feature phones as well as smartphones. Plus, AWS provides a timely response should any issues with the service arise and we are online 99.9% of the time.”

This allows the company to deploy its products speedily at the same time train accurate machine learning models that meet industry standards for image classification.

“We are enabling local farmers to yield better produce via our AI and machine learning capabilities, which are accessible via AWS,” Simbeye concludes proudly. “Our vision, is to improve food security and make a positive impact in solving farmers’ challenges with access to agricultural information in Africa.”