Back in 2016, I joined the University of Toronto’s student aerospace team with my friends Ayaan and Shayaan Haider. We’d build our own makeshift drones then make them square off in silly competitions, like seeing whose model could fly eggs a few kilometres without cracking them, or deliver cans of beer to frat houses without shaking them up. We noticed our fellow students were ordering items from Amazon Prime that they could easily get from local corner stores, so we created a delivery app, programming our drones to zip around buildings and other city structures to drop off orders.

Creating flight paths that drones can navigate autonomously isn’t as simple as plotting routes on Google Maps. When you look up, you may see clouds, telephone wires, maybe a few birds. But from a robotics standpoint, all kinds of invisible factors influence the accuracy of a drone’s route, like its altitude, electromagnetic and radio-frequency interference, GPS and telecom signals and, yes, weather. In 2018, the Haiders and I started capturing these datasets using sensor-equipped drones, bikes and other modes of transportation to map out cities like Vancouver, Toronto, Ottawa, Calgary and parts of Montreal.

That’s the origin story of AirMatrix. Today, it’s a master-control software that helps pilot autonomous drones safely, simultaneously and compliantly using all of the aforementioned data, updated in real time with an assist from AI. It’s now being used by the Canadian and Ukrainian militaries, as well as 30 American organizations, including the Federal Aviation Administration, or FAA. Soon, we’ll deploy in Riyadh.

Skyways are quickly going to become essential infrastructure for drones, just as roads are for cars. Between 2019 and 2025, Transport Canada certified more than 107,000 drone pilots, and our domestic........

© Macleans