Gunho Sohn
An innovative project that blends artificial intelligence (AI) with train safety has netted Lassonde School of Engineering Associate Professor Gunho Sohn and project partner Thales Canada a coveted second place in the Professional Engineers of Ontario (PEO) Project of the Year competition.
The Visual AI for Autonomous Train: The Ontario Train Autonomy Collaboration (OnTRAC) Project, an Ontario’s Autonomous Vehicle Innovation Network (AVIN) Stream 2 Grant-funded initiative between Sohn of the Department of Earth & Space Science & Engineering, and Thales Canada was submitted to the competition for the 2020 PEO York Chapter Engineering Research Project of the Year Award.
With the OnTRAC project, Sohn is advancing autonomous rail transport through the development of an AI-based computer vision system that will replace existing train control systems. While automated trains are prevalent worldwide, autonomous trains represent the next evolution in rail transport. Using Sohn's system, if a train or vehicle were to become compromised or encounter a safety hazard, human intervention would not be needed to allow for safe handling of the situation.
Professor Gunho Sohn (right) receiving his award outside the PEO York Chapter with two PEO York representatives
For autonomous trains to become a true successor to automated trains or vehicles, these systems need to be able to perceive their surroundings and make “human-like” decisions based on these perceptions. This is known as autonomous vehicle recovery (AVR), where the vehicle senses an obstacle and reacts appropriately.
The global autonomous trains market is anticipated to reach $16.59 billion by 2030. The outcomes of this project form the foundation to commercialize the primary component of Level 1 autonomous trains as part of Thales Canada’s next-generation autonomous rail signalling and control products. This project represents a major stepping stone towards Canadian leadership in global autonomous systems.
Images of the 16.7-kilometre section of rail on the York-Durham Heritage Railway (YDHR) in Durham, where Sohn, his team and Thales are testing their obstacle detection, classification and tracking (ODCT) perception functions via a custom sensor cart
Canada’s rail transportation system is the fifth largest in the world, covering more than 46,000 kilometres and transporting over 75 million people annually. This work can increase the efficiency and safety of future rail transportation systems in the country and worldwide.
"Professional Engineers Ontario York Chapter is proud to recognize engineering excellence and technical innovation through our annual Engineering Project of the Year Awards competition,” said EIT, Awards and Recognition Director Shengdi (Sharon) Chen of the PEO York Chapter. “The competition connects companies and institutions to their local engineering community and potential community partners, and provides a wide public forum for competitors, including Professor Gunho Sohn's team at York University, to showcase their technical expertise and engineering successes.
The Engineering Research Project of the Year Award is a brand new award launched in 2020, making the York University team one of the first to receive honours for it. We would like to congratulate the York University team on its second-place win and encourage everyone within the engineering community to apply for the 2021 award competition,” added Chen.
This project is supported by the Autonomous Vehicle Innovation Network – AV R&D Partnership Fund – Stream 2, administrated by the Ontario Centres of Excellence, which was awarded to Professor Sohn and Thales in 2019.
Sohn received the second-place commendation and trophy at the PEO York Awards ceremony on April 21. To learn more about the project, visit the "Learn more about the project" section at gunhosohn.me/category/ontrac.