Ottawa company teaching autonomous vehicles to ‘see’ snow, drive in bad weather

When a significant snowstorm hits Ottawa, most residents retreat indoors, griping concerning the weather and the heaping piles of snow they’ll soon need to shovel.

But for Fahed Hassanat and his team at Sensor Cortek, a giant dumping of snow is cause for excitement.

“The badder, the higher,” said the Ottawa-based software company’s chief operating officer and head of engineering.

“Nowadays you don’t get those heavy snowfalls as steadily as before, so every time they occur, we just rush rush, rush.”

Drivers for Sensor Cortek get behind the wheel of a automotive covered in sensors from bumper to rooftop and hit the road so as to solve considered one of Canada’s biggest roadblocks to autonomous vehicle adoption: snow.

Snow may be hard to tell apart for sensors, which are sometimes obscured and confused by bad weather, to detect, making it even tougher to coach self-driving vehicle software and algorithms.

And it’s not only snow that poses an issue.

“You’ve got the snow, you have got the rain, you have got the fog, you have got the dust,” said Hassanat.

To show a automotive to deal with whatever Mother Nature throws its way, Sensor Cortek outfits vehicles with laser imaging, detection and ranging (Lidar) sensors, radar, cameras and advanced GPS systems.

Lidar sensors emit laser beams and capture the reflections of those beams from the environment to create clouds of points in 3D space that provide details about location of objects in that space. Radar relies on electromagnetic waves to capture details about surroundings.

“(Radar) is a really complex sensor … but what’s so good concerning the sensor is that it will probably be covered in mud,” Hassanat said. “You’ll be able to put it any weather condition and they’ll still operate.”

Lidar and cameras are based on having line of sight, making them vulnerable to any obstruction between the sensor and the item it must detect, whereas radar doesn’t have to “see” an object to detect it.

They generate about 10 gigabytes of knowledge for each minute Sensor Cortek uses them, often at Area X.O, a personal, 1,850-acre site in Ottawa with a 16-kilometre track, where equipment starting from farming machinery and military tanks to emergency vehicles may be tested.

The work to get driverless cars able to handle any weather is crucial because even the best automotive manoeuvers may be complicated by bad driving conditions and make for “terrible, dangerous scenarios,” said William Melek, director of the University of Waterloo’s RoboHub, a robotics and automation research hub.

“Take into consideration autonomous cars attempting to make right in a snowstorm in a busy intersection … and you have got cyclists, you have got can hardly see street signs or markings,” he said.

“The data that’s coming out of your sensors or your data is totally unreliable, just because the software is unable to process occluded images or cluttered image with lots of noise.”

For Sensor Cortek, no two drives are alike because its automotive encounters different numbers of pedestrians moving in various ways, travels at changing speeds and finds itself in shifting weather conditions every time.

Sometimes Hassanat returns with a number of inches of snow sitting on the sensors. Other times they’re bare, but he’s encountered winds or blowing snow and debris that affected sight lines.

After a drive, Sensor Cortek uses the captured data to coach artificial neural network models, which later may be paired with the sensors to detect road users in all weather conditions and enable vehicles to make safer decisions.

The tip goal for Sensor Cortek is making a deep neural network and AI-based perception systems that could make autonomous vehicles — and anything needing sensors of the identical nature — “see” higher in all visibility conditions and ultimately, operate safety.

Auto and tech firms developing autonomous cars have encountered difficulty in extreme weather conditions, Google owner Alphabet Inc. admitting its own driverless vehicle project had found snow a struggle way back to 2015.

“It seems in Mountain View it doesn’t snow,” the corporate’s self-driving automotive project director Chris Urmson reportedly said of the automotive’s California testing location on the annual Automotive News World Congress conference in Detroit that yr.

Since then, there’s been a gradual flow of driverless automotive projects being abandoned for reasons starting from cost profit analyses to safety.

Automakers Ford Motor Co. and Volkswagen AG abandoned Argo AI, their autonomous vehicle company, last October, saying they don’t see a path to profitability for the project.

Uber Technologies Inc. sold its self-driving automotive division, which had staff in Toronto, to autonomous vehicle startup Aurora in 2020, following an incident where considered one of its test vehicles struck and killed a pedestrian in Arizona.

Melek believes within the short-term the businesses that keep at it’ll develop their cars to the purpose where some self-driving capabilities can be possible with human intervention, but he expects it to take many more years for a fully-autonomous vehicle to hit the road — and Canadians to trust it.

“I wish I can have a more optimistic outlook, but I’d say in my very own personal assessment — and I may very well be improper — we’re perhaps 15, 20 years out.”

This report by The Canadian Press was first published March 10, 2023.