Iris Automation raises $13 million for visual drone object avoidance tech

It's only a matter of time now before drones become a key component of everyday logistics infrastructure, but there are still significant barriers between where we are today and that future - particularly when it comes to regulation. Iris Automation is developing computer vision products that can help simplify the regulatory challenges involved in setting standards for pilotless flight, thanks to its detect-and-avoid technology that can run using a wide range of camera hardware. The company has raised a $13 million Series B funding round to improve and extend its tech, and to help provide demonstrations of its efficacy in partnership with regulators.

I spoke to Iris Automation CEO Jon Damush, and Iris Automation investor Tess Hatch, VP at Bessemer Venture Partners, about the round and the startup's progress and goals. Damush, who took over as CEO earlier this year, talked about his experience at Boeing, his personal experience as a pilot, and the impact on aviation of the advent of small, cheap and readily accessible electric motors, batteries and powerful computing modules, which have set the stage for an explosion in the commercial UAV industry.

You've now shattered some of the barriers that have been in aerospace for the past 50 years, because you're starting to really democratize the tools of production that allow people to make things that fly much easier than they could before," Damush told me. So with that, and the ability to take a human out of the cockpit, comes some interesting challenges - none more so than the regulatory environment."

The U.S. Federal Aviation Administration (FAA), and most airspace regulators around the world, essentially break regulations around commercial flight down into two spheres, Damush explains. The first is around operations - what are you going to do while in flight, and are you doing that the right way. The second, however, is about the pilot, and that's a much trickier thing to adapt to pilotless aircraft.

One of the biggest challenges is the part of the regulations called 91.113b, and what that part of the regs states is that given weather conditions that permit, it's the pilot on the airplane that has the ultimate responsibility to see and avoid other aircraft," That's not a separation standard that says you've got to be three miles away, or five miles away or a mile away - that is a last line of defense, that is a safety net, so that when all the other mitigations that lead to a safe flight from A to B fail, the pilot is there to make sure you don't collide into somebody."

Iris comes in here, with an optical camera-based obstacle avoidance system that uses computer vision to effectively replace this last line of defence when there isn't a pilot to do so. And what this unlocks is a key limiting factor in today's commercial drone regulatory environment: The ability to fly aircraft beyond visual line of sight. All that means is that drones can operate without having to guarantee that an operator has eyes on them at all times. When you first hear that, you imagine that this factors in mostly to long-distance flight, but Damush points out that it's actually more about volume - removing the constraints of having to keep a drone within visual line of sight at all times means you can go from having one operator per drone, to one operator managing a fleet of drones, which is when the economies of scale of commercial drone transportation really start to make sense.

Iris has made progress towards making this a reality, working with the FAA this year as part of its integrated pilot program to demonstrate the system in two different use cases. It also released the second version of its Casia system, which can handle significantly longer range object detection. Hatch pointed out that these were key reasons why Bessemer upped its stake with this follow-on investment, and when I asked if COVID-19 has had any impact on industry appetite or confidence in the commercial drone market, she said that has been a significant factor, and it's also changing the nature of the industry.

The two largest industries [right now] are agriculture and public safety enforcement," Hatch told me. And public safety enforcement was not one of those last year, it was agriculture, construction and energy. That's definitely become a really important vertical for the drone industry - one could imagine someone having a heart attack or an allergic reaction, an ambulance takes on average 14 minutes to get to that person, when a drone can be dispatched and deliver an AED or an epi pen within minutes, saving that person's life. So I really hope that tailwind continues post COVID."

This Series B round includes investment from Bee Partners, OCA Ventures, and new strategic investors Sony Innovation Fund and Verizon Ventures (disclosure: TechCrunch is owned by Verizon Media Group, though we have no involvement, direct or otherwise, with their venture arm). Damush pointed out that Sony provides great potential strategic value because it develops so much of the imaging sensor stack used in the drone industry, and Sony also develops drones itself. For its part, Verizon offers key partner potential on the connectivity front, which is invaluable for managing large-scale drone operations.