Machine Learning Edge Computing on Cassava Disease Models

A Machine Learning Research project conducted by a team of 2 developers and 1 project manager, and supported by Borealis AI ML engineer mentors. Borealis AI is the AI research pillar of the Royal Bank of Canada (RBC).

This research targeted the lack of technology in rural areas of our team’s home country, the Philippines, and how AI could be integrated into small devices (Edge Computing) to solve agricultural issues. Our team trained a Machine Learning model that uses image classification in detecting early stages of diseases among cassava plants; this can then be reused for other plants.

💡 The basics of picking an appropriate dataset (from Kaggle) and identifying its limitations
‍
💡 How to train an ML model using FastAI and what metrics to look at as a success criteria
‍
💡 How to plan, delegate tasks, and set goals for a Machine Learning research project
‍
💡 How to decide on what problem we want to tackle, identify the requirements, & decide between technological tradeoffs for a solution
‍
💡 Presenting a technical project to business stakeholders