Patrick Pilarski’s first passion was science fiction. Books like William Gibson’s Neuromancer sparked a childhood curiosity with artificial intelligence and cybernetic augmentation, which may have tangentially inspired his second passion: physiology.
He discovered this in a sports medicine program in his Stony Plain high school. The teenager became fascinated by the nature of the human body — and how to help it recover from injury and illness. “Back in the ’90s I was thinking, ‘Wouldn’t it be cool to build robotics to augment human abilities?” he says. “Fast forward … and I’m doing that exact thing.”
Since 2014, Pilarski has co-led the University of Alberta’s BLINC Lab — or Bionic Limbs for Improved Natural Control — where they research, design and test cutting edge bionic limbs for people who need them. Some patient research partners, like Chris Neilson, who lost his arm in a mining accident, have worked with the lab for over a decade. Along with co-leader Dr. Jacqueline Hebert, a clinician at the Glenrose Rehabilitation Hospital and professor of medicine at the University of Alberta, Pilarski works with research participants to improve the lab’s technology, while also collaborating with student researchers to develop and troubleshoot the technologies.
Though prosthetics have existed since at least ancient Egypt, and robotic prostheses have been commonly used since the 1960s, BLINC prosthetics use artificial intelligence to behave more like a living limb, intuitively adapting to the rhythms of life. “Since time immemorial, humans have integrated tools into their daily lives,” says Pilarski, likening their advanced tech to hammers and arrows. “And now, because of this rapid technological change in sensors, actuators, computing technologies, machine learning technologies, something different has happened. We can now think about our prostheses as collaborators in living lives.”
Since its foundation in 2014, the BLINC lab has explored some wackier possibilities, including an extendable, “go-go gadget” forearm prostheses, and a wearable third arm that comes out of the chest. As silly (or nightmarish) as these sound, their limitless imagination has led the lab to innovative places. Equipped with everything from 3D printers and augmented reality gear — and staffed by researchers from diverse fields — the lab is transforming what is possible with prosthetics.
But the real game-changer of the past few years is their integration of AI reinforcement learning, training software through trial and error in a dynamic environment, which the lab harnesses to make prosthetics more intuitive.
Movements as simple as passing salt around a dinner table are quite complex on a muscular level. To replicate that with a typical transhumeral (above elbow) prosthetic would require the user to closely concentrate on the maneuver as they manually change the direction of their prosthetic. Using reinforcement learning techniques, the prostheses “learn” to coordinate their motors and provide more lifelike and natural movements.
While these devices present exciting possibilities for amputees, they’re not without baggage, admits Pilarski. People working in tech start-ups can sometimes create “solutions looking for problems.” It can be tempting to just build something cool (like a chest-mounted third arm, perhaps) without considering standards of care, or customer needs. But, Pilarski says, these integrations prioritize quality of life by restoring natural movement.
Integrating machine learning and robotics also gets to the heart of what makes a human human. If a robotic limb uses artificial intelligence, then how do we preserve that individual’s autonomy? Where does the person end, and the robot begin? The lab is interested in these philosophical questions, and in fact has philosophy students in the lab, alongside kinesiologists and biomedical engineers. It’s a testament to BLINC’s interdisciplinary values and vision. “We don’t force all the computing sciences to sit on one side and then all mechanical engineers over here and medical residents only come in every third day,” Pilarski explains. “We deliberately build our projects and our space so that people are near each other, they’re working with each other — they can create, and I like to say they can constructively eavesdrop.”
Eavesdropping, he believes, forces interdisciplinary researchers to consider, and reconsider, every angle of their innovations. Some of the work might be dreaming up outlandish ideas, but what’s made in the lab must always serve amputees.
This article appears in the May 2025 issue of Edify
