Safari during Data Collection - Data Collection.

Through my PhD studies, Uganda’s Mango Farmers will benefit from the use of Artificial Intelligence-enabled environments and Machine Learning approach to fruit fly control and damage detection.

In the latest developments, the IoT-ra Lab based at the College of Computing and Information Sciences (CoCIS) at Makerere University is implementing the Adaptive Environmental Monitoring Network for East Africa (AdEMNEA) Project. The research study uses Technology-based Image Processing Techniques to help in early detection of Pest Eggs & Larvae i.e. Hyper-Spectral Imaging.

Safari uses Hyper-Spectral Imaging Technique on a Mango Farm.

Safari Yonasi, a PhD student at CoCIS focuses on building Machine-Learning models that will help in;

• Estimating and quantifying Damages caused by Fruit Flies,

• Determining the magnitude of fruit fly Infestation in relation to Climatic Conditions (Variations in Quantity of Fruit Flies)
• Building risk maps to visualise – show how Pests Damage the different parts of Farms, identify various Pest Species on the Farms.

“These models leverage cutting-edge artificial intelligence (AI) methods to analyse various features and accurately assess the extent of damage inflicted on the fruit. By utilising high-resolution images and artificial intelligence techniques, the models can detect various damages to the fruit's texture, which offers valuable insights to orchard management,” Safari emphasized.

During the past few weeks, the tech-savvy student has been engaged in the field, collecting image data-sets for mango fruits across different varieties. He is positive to reap maximum results given the timely nature of his research and the relevance the research has to mango farmers.

Dr. Rose Nakasi, a PhD researcher in Machine Learning says Safari’s research work draws insights in timely fruit damage predictions and also leverage the same for similar agricultural problems for effective interventions.

Dr. Rose Nakasi, a Computer Scientist on the AdEMNEA Project and one of Safari's Supervisors.

“Artificial Intelligence has the capability to learn and detect particular agricultural phenomena like fruit damage and fruit fly abundance prediction with more precision, while also leveraging on other datasets such as weather data in the areas where similar variables exist, to further improve the detection perfomance.” The data scientist stated.