News & Events

Opening a door isn't a simple job for robots. For instance, the activity requires a robot to spot and recognize the handle, grasp it and then determine the necessary force to pull (or push) the door open. It may be a multi-step process, but controlling a robot, and having it complete this task reliably, is critical to teaching it to manage more complicated ones in the future, says researcher Guangjun Liu.

Liu, a professor in the department of aerospace engineering, holds a Tier II Canada Research Chair (CRC) in Control Systems and Robotics. Renewed for a second five-year term in 2011, the prestigious chair has enabled Liu to focus on research that helps maintain Canada's position as an international leader in robotics and aerospace engineering.

"The renewal of Dr. Liu's Canada Research Chair in Control Systems and Robotics is a testament to the quality of his work," says Wendy Cukier, vice-president, research and innovation. "Guangjun's developments in robotics have multiple applications and have a dramatic impact while offering sustainable solutions for both economic and social benefit. It is a great example of how research at Ryerson is helping drive innovation."

During Liu's previous CRC term, he also worked on aircraft environmental control systems, making them more efficient and thereby reducing drag and fuel consumption. This time around, though, Liu is focused on developing autonomous robots for unstructured environments such as space, high-level surveillance situations and health-care settings. Liu is well-versed on this and other topics in his discipline; he has authored or co-authored a total of 33 journal papers within the last five years.

"Robots have long been used successfully in the structured environments such as assembly lines," says Liu. "The trend now is to have them work in unstructured environments - exploring space, providing security in airports or helping to care for patients. But to do all that, robots must know how to adapt to their surroundings."

A key component of that adaptability is something called multi-mode control. As Liu and his research team are demonstrating in his lab, it's not enough to teach robots how to open only one kind of door. In a world of glass doors, steel doors and even office doors, a robot must understand how to handle each type of opening - and use the knowledge gained in one situation to complete increasingly complex tasks down the road. Those future activities could involve everything from delivering a glass of water to a hospital patient to picking up rock samples on the surface of Mars.

For that reason, it's important that robots also possess the ability to roam different areas - and have equipment onboard (e.g., a small manipulator arm) to accomplish tasks along the way. This mobility, though, comes with its own set of problems, according to Liu. "In structured environments, robots are tied to a power source. But in unstructured environments, robots need reliable sources of power that can travel with them, such as batteries."

Equally important, robots must constantly monitor their work performance and their ability to efficiently manage power. They must also be able to detect technical difficulties, such as slower movements or processing times, as soon as they arise. Recognizing these problems before they become crises - and alerting operators that repairs are needed - is crucial, says Liu. "This optimizes the robot's performance and ensures that it's ready to work whenever needed."

In addition to the Canada Research Chairs program, Liu's research has received funding from the Canada Foundation for Innovation, the Canadian Space Agency, the Natural Sciences and Engineering Research Council of Canada (NSERC), as well as Honeywell Aerospace, MPB Communications Inc., Engineering Services Inc. and PPIC (recently bought by Pure Technologies Ltd.).