Boeing Vancouver hopes to transform aircraft maintenance technology with augmented reality
The company joined forces with SFU and U.S. game develop Unity to develop a potentially revolutionary app
You have to admire the way Jack Hsu explains the principles of Boeing’s augmented reality technology to a complete Luddite. I didn’t think I was one of those, but listening to Hsu—a senior manager at Boeing Vancouver’s Richmond facility—wax poetic about the aircraft manufacturer’s innovations in maintenance, it quickly becomes obvious.
With support from Canada’s Digital Technology Supercluster, Hsu’s team at Boeing Vancouver partnered with SFU and San Francisco–based video game developer Unity to create an app that makes aircraft maintenance easier by using AR to visualize problem areas.
“Today, every dent and buckle or defect in an aircraft is stored in a database somewhere and tracked over time,” Hsu says. “Typically the way it’s done today is through a logbook—a mechanic will walk around, identify where the dent or buckle is and note that it’s three inches down from the fifth window on the left hand side by the door, for example.”
Hsu and Boeing Vancouver, which has about 200 employees across its two offices in Richmond and Yaletown, figured there had to be a better way to identify such imperfections.
The three organizations teamed up to create a specialized AR engine to easily document the places where maintenance is needed and isolate them.
Hsu compares the technology to the mobile game Pokemon Go. “It’s like seeing little Pokemon on the plane,” he says. “As you walk around the plane, the Pokemon stays stuck in the same location. It’s a more intuitive way of doing it, and you can do all sorts of things with augmented reality—can colour the defects that already exist, look for new damage, create metadata associated with a new piece of damage.” Using AR headsets, mechanics can quickly assess and then fix the damaged areas.
Boeing Vancouver, SFU and Unity tried out their project on an old Douglas DC-3 at Langley Regional Airport. Although the early results are “very promising,” according to Hsu, the project isn’t operational. “We ran out of time and dollars on the Supercluster project, but there’s a list of things we want to do to continue to improve the algorithm,” he says.
That includes building 3D models that would track everything under the skin of the aircraft as well, and even include guidance on how to fix a pump, for example.
The limitation right now is the devices, most of them created by a certain California tech giant, which don’t adapt AR particularly well to a large space like an aircraft hangar. But Hsu and his team are confident that they’re onto something and will be able to overcome that barrier as they keep the project going through new funding opportunities.
“Apple basically agreed; they know their tech has room to improve in those environments,” Hsu says. “So they're interested in our solution.”