Meet the robots of the future already hopping, slithering, and driving around U of T

While pop culture and science fiction have cemented the idea of humanoid robots in our collective imagination, the reality is most robots look nothing like us.

The physical AI revolution is underway, and Canadian research has paved the path.

Take the Hopcopter, which uses a combination of jumping and flying to move around more efficiently than its bipedal peers. Featuring four propellers atop a single pointed foot, the small bot looks more like an insect than a person. Its design allows it to operate for long periods by recycling energy, makes it adaptive to varying terrains, and lets it hop over large obstacles.

Song Li, a PhD student visiting the University of Toronto (U of T) from City University of Hong Kong, told BetaKit in an interview that these features make the LiDAR-equipped device particularly useful for mapping industrial areas.

Li demonstrated the Hopcopter’s capabilities manually using a PS5 controller. For a few minutes, it hopped around a gently confined space before crash-landing near our feet mid-conversation. “Sorry,” he said with a laugh, as he picked it up and made some adjustments.

The Hopcopter was one of the devices showcased as nearly 400 students, professors, and entrepreneurs seeking talent and potential partnerships gathered at U of T Mississauga for the school’s annual robotics conference on Tuesday and Wednesday, where they demonstrated some of the latest in robot technology.

A man adjusts a small robot. — The Hopcopter uses a combination of jumping and flying to move around more efficiently than its bipedal peers.

The physical AI revolution is underway, and Canadian research has paved the path. Industry leaders—like U of T professor and Waabi founder and CEO Raquel Urtasun, and Clearpath Robotics co-founder and Canadian Robotics Council chair Ryan Gariepy—think Canada has a chance to lead in the sector but is leaving opportunity on the table.

BetaKit caught up with U of T student researchers building robots for everything from lunar transport to surgery to autonomous vehicles. This group shed light on where some of that opportunity might lie and what next-generation, real-world robotics could actually look like.

At U of T’s Continuum Robotics Laboratory (CRL), researchers are developing “soft robots”: snake-like, jointless, tendon-driven robotics with a softer, more compliant, and adaptive touch than their rigid counterparts.

PhD student Jimmy Shentu demonstrated an elephant trunk-like robotic device’s ability to gently flick a light switch both on and off from various locations—a simple task designed to show the effectiveness of the CRL’s systems. Shentu told BetaKit the so-called “continuum” tech’s more human-like touch could make it better suited for human rehabilitation, accessing tight, hard-to-reach spaces, or certain types of surgery than hard robotic systems.

U of T’s Medical Computer Vision and Robotics Lab is focusing on a specific type of surgery: Twin-to-Twin Transfusion Syndrome (TTTS). TTTS is a rare and life-threatening condition treated with a technically-challenging procedure called fetoscopic laser photocoagulation that involves inserting a tiny camera and laser fibre into the womb to sever blood vessels in the shared placenta and allow each twin to develop separately.

A woman operates a robotic arm. — U of T undergraduate student Khushi Malik is helping to train a system to complete a rare surgery on its own using reinforcement learning.

Researchers at U of T and Mount Sinai Hospital developed a simulator that includes a disembodied, 3D-printed model of a maternal abdomen, a mock robotic arm surgical tool, and a virtual reality headset to help surgeons train for this procedure ahead of time.

U of T undergraduate student Khushi Malik is helping to train that system to complete the surgery on its own using reinforcement learning. She told BetaKit that more automation, coupled with doctor supervision, could potentially make this process safer and more effective.

Meanwhile, the university’s Autonomous Space Robotics Lab (ASRL) is shooting for the moon. PhD students Luka Antonyshyn and Anthony Beca are helping to develop a solution for lunar cargo transport. Their work has focused on getting a group of small, four-wheel robots to work together to carry larger items, and to independently move lighter loads, between landing sites and research bases on the moon. The idea is that multiple, smaller vehicles would be both cheaper and more versatile than a single, larger machine, they told BetaKit.

ASRL is doing this with Brampton’s MDA Space, as the latter explores a bid to build lunar utility vehicles for the Canadian Space Agency. BetaKit followed the ASRL’s three rovers on a planned path around the outside walkway. Barring some stops and starts thanks to minor intra-convoy collisions and oblivious attendees, the trio moved steadily as a single unit, keeping a consistent distance between each other as they looped back to their starting point.

A self-driving car. — U of T’s aUToronto operates a souped-up Chevrolet Volt that features a sophisticated autonomy and sensor stack developed over the past five years.

Back on Earth, U of T’s self-driving team, aUToronto, is fresh off another big win. The group’s souped-up Chevrolet Volt, which features a sophisticated autonomy and sensor stack developed and refined over the past five years, looks like a standard SUV from the outside, except for the large contraption on its roof.

That vehicle was on display yesterday after recently nabbing its eighth first-place overall finish in nine years at the General Motors-hosted international competition, as aUToronto outdid several other schools from across North America for top honours.

Next on aUToronto team principal Connor Wilson’s to-do list? Autonomously driving that car 18 kilometres down Toronto streets from U of T’s Downsview Park-based aerospace institute to its downtown St. George campus, something aUToronto aims to do within the year.

The U of T undergraduate student, who is also currently interning with Toronto’s Peripheral Labs, told BetaKit he is both nervous and excited by the challenge, noting that this trip will be “an order of magnitude harder than anything we’ve done so far.”

All images courtesy University of Toronto Mississauga. Photos by Nick Iwanyshyn.