Photo: iStock Photo: iStock

Drones, we’re told, are the future. The small flying bots can be used for a range of tasks, including filmmaking, industrial inspections, deliveries, and surveillance. But for the drone industry to really take flight (sorry), the bots need to be able to navigate on their own.

Perceptiv Labs is developing technology to do just that. The Waterloo-founded company is releasing a system later this year to give drones autonomy so they can zip around the skies with no human intervention. Co-founder and CEO Neil Mathew explains the company’s game plan. Why do drones need autonomy?

Neil Mathew: Drones are extremely hard to fly, and the potential for drones to create scalable data collection pools in various industries like filmmaking, industrial inspection, agriculture, security etc., is really high. The problem is the industry isn’t scaling fast enough because the way drones work right now, you need a manual pilot to come in and fly them. That’s fairly difficult to do.

The computer systems we’ve built let drones fly themselves and essentially do things that a pilot normally would, like be able to detect and track objects, fly without hitting anything, and fly without GPS. This is especially important in conditions where GPS isn’t reliable. So if you’re in a forest in Canada or near a building or bridge where GPS signals get wonky, it’s important that drones stay stable and safe.

Where did the idea come from?

We’re three engineers from Waterloo. We worked on computer vision and robotics for many years, and saw the potential for computer vision was huge. The initial applications were very limited. We were working on navigation using lasers and ultrasonic sensors, but the drone industry can’t support any of that because the equipment is too big and heavy for drones. They really need a lightweight, power-efficient sensing system that lets them fly themselves.
So we’ve been working around the constraints of drones and building extremely lightweight, powerful systems that can essentially do everything the Google self-driving car can, but at a fraction of the cost and in the air. Our whole system is about 200 grams and costs about $100 in terms of just the hardware. The idea is to make the hardware as cheap as possible, and make the software sophisticated.

Who’s your target market?

We’ve been working with professional filmmakers, who are the biggest users of drones right now. We’re building modules that are aftermarket upgrades for drones, so you can add these onto any drone and it’ll add a set of features to that drone to let filmmakers control their cameras in ways that they can’t manually. Controlling the drone to create footage is really challenging. Filmmakers have to hire people and they need to spend hours and hours doing multiple takes to get the right shot.

And how big is that market?

I was at the National Association of Broadcasters show last week and 97,000 filmmakers attended. Every filmmaker is adopting drones because it lets them move their cameras in new ways. If you use a jib or a steady cam, all you can really do is move the camera smoothly. If you want to start filming on the ground and fly up in the air, that requires a drone. In fact, drones can replace a lot of film equipment right now. A dolly can be completely replaced by a drone. We’re positioned to work with a whole spectrum of filmmakers, from professionals to amateurs.

Where are you at today in terms of development?

We’ve launched the tracking system for camera motion control, and so if you’re using a drone and camera gimbel, all a filmmaker has to do is put this system on their drone. That’s going to be shipping in the fall. We’ve sold it to a set of pilot customers, we’ve set up partnerships with professional filmmakers in Los Angeles, and we’re working with extremely specialized aerial cinematographers that work with some of the largest studios in L.A. We’ll work with them to refine that product, and learn the specifics filmmakers want, and tailor it. That’s going to happen over the summer, and in September, we’ll start shipping.

What about other potential customers?

For the first year, we’re going to be focusing on film. Right now, it’s not legal to fly a drone without line-of-sight, so the only way you can do it is by having the pilot stand within line-of-sight of the drone. But if a company wanted to inspect a field of 1,000 windmills, you can’t get a pilot to fly and inspect every single windmill. Ideally you want to get a team of drones to fly and inspect 1,000 windmills autonomously. Looking to the future, it’s easy to see that autonomy will be standard technology for drones. As soon as regulations clear the way for other industries, we’re in a good position to make the jump.

Why are you moving from Waterloo to Mountain View, California?

We’re essentially planning to move half the company down to Mountain View. Being able to access the network of customers, investors and other drone companies in the Valley is extremely valuable. Waterloo is great because we have a great network of engineers and our burn rate is low, but to really scale at a fast rate, there’s no substitute for the Valley.

What do you think of Mathew’s plans for Perceptiv Labs? Share your (constructive) thoughts and feedback in the comments below.


Have you launched a new business you’d like to put under the Business Plan microscope? Email us your pitch.

Loading comments, please wait.