Flying into the future: Yang Shi advances uncrewed vehicles on land, sea and air
January 23, 2025
A drone’s-eye view of Yang Shi would show us a person eyeing the future and imagining the next transformative idea.
“Every day,” he says, “we meet challenges. In response, my team and I come up with wild, audacious ideas. Sometimes they work.”
The UVic mechanical engineering professor’s ideas improve the performance of uncrewed vehicles in the air, on land, on and under the sea. He makes them more reliable, connected, communicative and efficient.
“A human operator can manipulate a drone in flight, sharing information between the controller and the drone,” Shi says. “And there’s always a time delay, however small, between sending a message and the machine receiving it and then responding to it. Ideally, we want the information sharing to be immediate, but even in the 5G era that’s not possible.”
However, for more than 20 years, Shi has continuously developed, expanded, applied and redeveloped a framework that accommodates, reduces and compensates for that communication delay.
Group performance
Inspired by a flock of birds, Shi knows that for a group of things to do a job — seagulls harrying an eagle, or drones monitoring an oil spill — the members must share information and adjust their individual and group performance based on that information.
“If we want drones to stay in formation to track an object, for example, we need consensus among them. So they don’t all have to communicate with every member of the group at once, each drone communicates with its immediate neighbours. This means that they’re all sharing only necessary information all the time.”
But there’s still going to be a time delay, of course, so Shi applies two important tools.
“We can teach the machines to predict and optimize,” Shi says. “We can send a prediction to the robot and it can use artificial intelligence to apply the prediction to future situations.”
Similarly optimization, performed in milliseconds, allows the machine to adapt to a new environment. A drone in flight might need to adjust to a gust of wind; if it has been given information in advance about what to do in that situation, the vehicle can predict what it needs to do to remain stable and can optimize its response: change its speed or angle of flight just enough, for example.
“We make use of AI-enabled control technologies to enhance robotics,” Shi says, but not only robotics. “We can also expand AI-enabled control to large-scale systems. If each flock of birds or drones is a system with multiple active agents, a smart grid or smart city is a multi-agent hierarchical system in which each element communicates with others within its layer, and between layers too.”
Global leadership
Shi was one of the pioneers in the world to tackle this ever-growing challenge, and his seminal theory, framework and algorithms have influenced engineers and engineering around the globe for more than two decades. He is a prolific researcher with a steady stream of awards and honours from academia and industry. Among many prestigious awards, he is a recipient of the 2024 Institute of Electrical and Electronics Engineers (IEEE) Canada Outstanding Engineer Award and the 2023 IEEE Dr.-Ing. Eugene Mittelmann Achievement Award, “for outstanding contributions to networked and distributed control, model predictive control, industrial cyber-physical systems and mechatronics applications.”
Part of his broad success is because of his innovative theorizing. Perhaps more is because of the myriad applications: one begets another, which translates to another and yet another.... People who nominate him for awards identify his vision and the innovations that enable adaptability, scalability, resiliency and usability of internet-linked systems. All of this has generated significant societal impacts.
One area of growing global importance is information security in industrial automation.
“Our system prevents external bad actors from hijacking information during a group flight, for example,” Shi says. “Each robot only shares part of the information or encrypted information, which helps preserve privacy.”
Remote robotic surgery is also feasible, given the speed and integrity of control, communication and adaptation that Shi’s work has made possible. And environmental monitoring and response in remote communities is yet another area that has huge and exciting potential for people and the planet.
With funding support from the Canada Foundation for Innovation, Shi’s laboratory team, including post-doctoral fellows and students from undergraduate to PhD, has built drones and robots with which they can test their algorithms. As well, their collaborators implement Shi and team’s theories on their own uncrewed vehicles, which validates their results in other settings and launches ideas for yet more applications.
Local impacts on climate and environment
For example, Shi is collaborating with Shift Coastal Technologies, a local company that uses ocean technology, expertise and industry relationships to support coastal First Nation communities’ planning, economic development and sustainability goals. Shi’s team is using uncrewed aerial vehicles and ocean-surface vessels to monitor environmental problems like oil spills.
He’s also applying the technology to electric vehicles. Remember those birds that need to coordinate the flock? Shi points out that autonomous vehicles too need cooperative controls.
“Also, we can manage power consumption by teaching them to choose an appropriate route using prediction and optimization.”
In a similar way, he’s working with the National Research Council and Transport Canada to optimize scheduling for vehicles using hydrogen fueling stations in order to reduce waiting times at the pump.
People and the next generation
“Probably the most important outcome of my research is highly qualified personnel [HQP],” Shi says. “They continue to develop new technology and to train other people.”
Indeed, of more than 150 HQP that Shi has mentored at UVic, almost half have faculty positions and the others are (often award-winning) senior engineers in major industry positions.
“Good teaching,” Shi says, “is closely related to the state of the art in research. We need to incorporate the research into the classroom, so I show them the latest technology in the world.”
His students appreciate his focus on giving them opportunities to learn critical thinking, systems approaches and the skills needed to innovate.
Mingxi Liu, now an associate professor at the University of Utah, was a student of Shi’s throughout his master’s and PhD programs. He credits Shi with helping him get an NSERC post-doctoral fellowship and a post-doctoral position at the University of California Berkeley, and the prestigious US National Science Foundation CAREER Award for early-career faculty members.
“Yang is creative, knowledgeable, and highly productive,” Liu says. “He has rich experience in working across traditional disciplinary boundaries and is the most rigorous researcher I have ever met. Overall, I do not have a sense of what the limits of Yang’s capabilities are, but I am confident we are nowhere near them.”
Another former student, now an engineer with Mercedes-Benz, adds, “Without his guidance and encouragement, I could not have found my dream job.”
“Through his remarkable supervision,” says fellow mechanical engineering professor Zuomin Dong, “Dr. Shi helped catalyze the careers of many gifted engineers and scholars in industry, research and academia.”
Every week, Shi meets with all his graduate students, at which one presents their own research while the rest provide feedback, plus he meets individually with each of them. He does this while maintaining such a prodigious and impactful research and publishing program of his own that he has been named Clarivate Highly Cited Researcher for ten years in a row.
Behind the publications and HQP, the fleets of drones and uncrewed vessels, the pursuit of energy efficiency and environmental responsibility, at heart Shi agrees with the maxim that there is no better application than a good theory.
Control theory, he says, plays an important role in supporting many applications and AI-enabled control theory will innovate ever more.
And in Shi’s world of cooperation and communication, truer words were never spoken.
Rachel Goldsworthy