Shahram Tafazoli
Founder & CEO
Motion Metrics
Dr. Shahram Tafazoli is the Founder & CEO of Motion Metrics International Corp., a market-leading Canadian technology company dedicated to improving mine safety and efficiency using artificial intelligence, computer vision, and cloud computing. Dr. Tafazoli conducted his thesis in the area of Robotics and Intelligent Systems at the University of British Columbia’s ECE Department and graduated in January 1997. Dr. Tafazoli is also an avid inventor holding several patents, an adjunct professor with UBC ECE, an angel investor in several promising Canadian tech startups, and an associate member of the Creative Destruction Lab in the University of Toronto.
09.10.2019 12:15 - MineDigital | Pushkin Hall
Particle Size Analysis in the Age of Artificial Intelligence: A Novel Approach
Mine-to-mill optimization begins with a clear understanding and ongoing assessment of blasting, crushing, and grinding operations. Modifying blast practices to optimize mill feed size can increase a mine’s throughput by up to 30% but establishing baseline performance and optimize feed size requires continual analysis at critical points in the size reduction process.
We developed a highly flexible and accurate approach to particle size analysis that uses a proprietary machine learning architecture, stereo imaging, and cloud computing to provide real-time results. This solution can be adapted for many applications in the pit and the plant, and is commercially available for shovels, conveyor belts, and in a portable format. An additional application for haul trucks is currently undergoing testing and will leverage the existing machine learning architecture to provide boulder detection to mitigate crusher downtime. Our solution continually learns and improves, guaranteeing performance in a broad range of environments.
10.10.2019 16:00 - “Mining Goes Digital” Conference - Session 4 | Main Hall
Automating Routine Operational Decisions with Artificial Intelligence
Over the last few decades, cameras and sensors have been deployed to replace or enhance traditional equipment monitoring methods. For example, modern solutions can detect broken shovel teeth, analyse particle size in real time, and monitor conveyor belts. However, humans are still needed to close the loop. Even as the accuracy of prediction improves, humans still need to correct algorithms where they fail and oversee the functioning of connected assets.
In the future, humans can be taken out of the loop. As the prediction accuracy of artificial intelligence improves, humans will no longer need to provide corrective action or supervision. For example, missing tooth detection solutions could automatically reroute haul trucks carrying broken teeth instead of notifying dispatch that an event has occurred. Automating the decision process will free up time for mine personnel to improve the safety and efficiency of operations worldwide.