Drone air vehicles, a swimming robot, an app that helps the blind see, an artificial pancreas, a wheelchair that drives itself. These are some examples of the cutting-edge research being done at the Centre for Intelligent Machines (CIM). With 20 professors and 150 graduate students, CIM draws students and faculty from around the world to McGill.
For a world-renowned hub of artificial intelligence and robotic research, CIM is a bit hard to locate on campus. As you get off the elevator on the 4th floor of the McConnell Engineering Building you may notice a small sign directing you toward CIM. It is a particularly modest sign for a multidiscliplinary research centre with a pedigree as rich as CIM’s.
“Everyone in the field has heard of CIM,” says Carmen Au, a graduate of CIM who now works as a software developer in Silicon Valley. “Not only that, the CIM was very welcoming to women. We were judged only on our work. I relished the interdisciplinary and collaborative atmosphere at CIM when I was doing my PhD there. In Professor Jim Clark’s lab I was exposed to so much work being done by others. I didn’t realize how amazing it was when I was there, how few silos there were, until I left CIM.”
Jim Clark, CIM Director, says Au’s impressions of the Centre’s collegiality are in line with the spirit in which it was established. “CIM was created to foster interaction between researchers and facilitate collaboration. CIM professors regularly co-supervise graduate students and serve on other members’ student committees.,” says Clark. “There are many large funded collaborative projects: the NSERC Canadian Field Robotics Network, the AQUA robotics project, the NSERC Create program in Medical Image Analysis, and the APC Electric Drivetrain project, bringing together multiple professors and students solving common problems. This is made easier by CIM’s shared infrastructure, with laboratory spaces serving multiple research groups, and a common computing environment.”
Back in 1985, Martin Levine was one of the first people working on and teaching Computer Vision – then back Pattern Recognition.
He also helped found CIM back in 1985. “Just like humans and animals, robots need to see what they are doing to function properly,” says Levine, now Emeritus Professor in CIM’s Visual Surveillance Group. “We submitted a grant application to NSERC and, amazingly, were funded for the project. [With that money] we created CIM in 1985, and eventually expanded to what we are now.”
AI is a focal point for CIM researchers. “Fully half of CIM’s professors are engaged in some form of Artificial Intelligence research, including AI for robotic vision, and AI for medical image analysis,” says Clark. “They include Professors Joëlle Pineau and Greg Dudek, head of the NSERC Canadian Field Robotics Network; Paul Kry; Tal Arbel; and Kaleem Siddiqi, among others.
Clark says CIM is an important institution in the Quebec, Canadian, and world-wide context. Its collaborative approach and diverse teaching staff allows it to develop ideas and confront issues that few other centres can.
“CIM is one of the few research centres in North America, and perhaps the only one in Canada, that brings together such a diverse range of researchers to work in the broad expanse of intelligent machines,” he says.