2026 JOHN OGILVIE RESEARCH INNOVATION AWARD

“Automated Smart Imaging System for Real-Time Canola Seed Quality Detection”

Started as a small seed grant of $9,984 funded by the University Research Grant Program (URGP) of the University of Manitoba, Dr. Erkinbaev had proposed the development of a smart imaging sensor for real-time determination of canola seed quality. He was driven by the idea of moving away from the visual inspection of canola seed by trained personnel, which was a labour-intensive, subjective and time-consuming process. Later, this research has grown into a larger initiative with several funded research and infrastructure projects, which aimed to help with quality control of canola from farm-to-fork and provide an intelligent tool for better and more efficient decision making in the canola processing industry. With a team of only one MSc student (Mr. Angshuman Thakuria), a hypothesis was tested by developing detection algorithm of damaged seeds based on image and video recordings coupled with an edge device. They published their research findings and results on achieving higher accuracy, less computational models and faster speed in a peer-reviewed journal in the area of smart agricultural technology. Inspired by the preliminary results from this seed grant, they were excited to expand the project with a focus on industrial application, which resulted in securing the highly competitive Proof-of-Concept Grant ($100,000) from the Research Manitoba and the Manitoba Government ($129,000) for the development of a smart imaging system for real-time canola quality prediction (Mr. Thirukumaran Ramesh). Innovative nature of the project led to expanding the objectives from monitoring of incoming raw canola seeds to prediction of intermediate and final canola oil and meal products resulted in securing two capacity building grants ($316,800) from the Manitoba Sustainable Canadian Agricultural Partnership, Manitoba Government, for development of an online quality control system for canola processing: from single kernel characterization to oil and meal products.