Lori DeMatteis, Iris Automation: “in the near future, many industries will utilize drones in scenarios that were previously impossible”
The recent advancements in technology allow more and more everyday tasks to be completely automated and fulfilled by drones.
While drone technology is great for a variety of purposes from entertainment to deliveries or surveillance, experts warn that with the growing rates of uncrewed machines in the air, issues like collision or unauthorized flights can become a problem.
To discuss how the power of AI can be used to ensure safe flights, both for crewed and uncrewed flight, the team at cybernews sat down with Lori DeMatteis, Global Vice President of Customer Success, Sales, and Marketing at Iris Automation.
How did Iris come about? What has the journey been like since your launch?
Iris Automation was founded in 2015 and came about because flying drones was limited to flying within a visual line of sight. A pilot, typically on the ground, had to be able to see the drone at all times, which limits its flying range to about a mile, and to what could be tracked directly in front of them.
With major developments around AI and machine learning, our founders saw the possibility of harnessing this technology to build a detect and avoid solution similar but better than the human pilot’s eyes – preventing collisions with other flying objects in a 360 field of view. By automating detection we vastly improve the ability to see and identify other objects. This resulted in Casia, our detect and avoid (DAA) solution which has grown into multiple products and software releases. We enable drones to fly beyond the visual line of sight (BVLOS) but also provide an additional layer of safety for crewed flights.
We’ve worked closely with the aviation industry and government regulators such as the FAA here in the US, Transport Canada, EASA (the EU’s aviation safety regulator), and other international bodies to help progress regulations around autonomous flight. We are part of the FAA’s BEYOND Program, working closely with the City of Reno in Nevada, for example, to test the deployment of drones in search and rescue operations.
We have partnerships with major aviation companies such as Embraer and Becker Avionics who want to investigate autonomous flight or add Casia to crewed flights to provide an added layer of safety.
Can you introduce us to the Casia solution? What are its key features?
Iris Automation pioneered Casia I and Casia X, the first electro-optical on-board, autonomous commercial collision avoidance safety system for both uncrewed and crewed aircraft systems (UAS). Casia G leverages the same AI and computer vision technology, but is ground-based and stationary, to detect aircraft as they approach your UAS operation. Casia G is useful when a drone operator wants to limit additional payload, freeing it up for other applications or cargo.
Casia uses Iris Automation’s patented detect and avoid technology to detect, alert, identify and avoid non-cooperative aircraft be they crewed or uncrewed. For operators to fly autonomously over distances, onboard sensors like Casia X or ground-based sensors like Casia G, combined with carefully designed concepts of operation, provide cost-effective and scalable solutions.
Casia’s reliable collision avoidance consists of these features:
Detect
• 2000m to 2400m detection range for non-cooperative intruder aircraft depending on application
• Compatible with ADS-B for cooperative traffic and other DAA sensors
• Full optical situational awareness
Avoid
• Real-time aircraft classification through Casia’s proprietary algorithms
• Intelligent autonomous avoidance by determining the location, heading, and speed of intruder aircraft
• Built to meet emerging regulatory DAA performance standards
• Powered by cutting edge computer vision technology
• Machine learning and computer vision architecture for industry-leading performance
• Tools to manage and replay flight data
• Configurable device setting to customize behavior for specific mission profiles
• Ongoing software updates improve performance and capabilities over time.
What would you consider to be the most pressing challenges surrounding drone use nowadays?
Many organizations involved in large-scale inspections and surveillance, whether of pipelines or ports, wildlife sanctuaries, airports, and cities, are using (expensive) helicopters, inspecting by foot or truck, or are simply not able to inspect an area. Using drones is a game-changer. It can lower environmental impacts as compared to traditional options, provides a more cost-effective and efficient solution, and provides critical services to remote locations such as medicine and specialized healthcare.
But today there is a major effort required by any commercial drone operation planning to cover significant distances due to the BVLOS requirement set by global aviation regulators. Without a BVLOS waiver, they must land their drone, move to a new base of operation, and relaunch to cover distances of just over one mile.
Meeting a country’s aviation regulations for uncrewed, long-distance (BVLOS) flights is the number one challenge any operator must meet. Most countries now offer waivers if an operator meets certain criteria for their operations but they are stringent, and designed to remove the risk of collision. They require comprehensive documentation covering the mission, the aircraft, and any additional equipment such as communication links and DAA technologies.
Organizations planning to operate large-scale drone services must therefore aim to standardize their unmanned system platform with one manufacturer, for example, and provide an additional layer of safety like our Casia detect and avoid solution to meet these regulatory requirements. This enables repeatability and scalability as you expand operations.
How did the recent global events affect your field of work? Were there any new challenges you had to adapt to?
It limited our ability to host physical demonstrations at our flight operations center, but overall the pandemic demonstrated the demand for greater flexibility around autonomous delivery methods, remote, uncrewed flights and inspections, and the use of AI and machine learning to automate previously manual activities.
Due to the pandemic, many structural inspections were forced to be performed by drones – such as bridge, road, building, and pipeline inspections, and deliveries of packages and medical supplies. This certainly created greater awareness and adoption of the work we do.
Lastly, we are seeing supply chain issues across all the manufacturers we work with. This has forced them to think creatively, and have multiple sources or options for parts, or substitute parts or requirements so that there is no one single point of stoppage.
What other industries or company processes do you think would greatly improve by implementing automation?
As autonomous flight becomes more mainstream we’ll see commercial drone operations become more widely used either through a service provider or by an organization running its own drone operations. We’ll see more widespread use of delivery services and a wide range of industries will utilize drones in cases where they were previously impossible, impractical, or cost-prohibitive. We will start to see piloted single-passenger taxis and eventually autonomous uncrewed aircraft.
We also see an opportunity to dramatically help isolated communities which typically suffer from a lack of resources and access. For example, autonomous flights will make it feasible to transport people to major hospitals, typically in urban areas, which would have been cost-prohibitive with traditional flights (and frequently an airport isn’t even nearby) or long drives, if the roads are passable.
Drones can also enable cost-effective deliveries not only of medicine and health supplies but also books and educational resources to isolated areas. They can survey areas during natural disasters fast and cost-effectively.
Besides providing onboard collision protection, you also monitor the airspace for intruder drones. What drone threats do you find the most concerning nowadays?
General aviation pilots are concerned about colliding with uncooperative drones – it’s easy to buy a consumer drone online and start flying it, without any permit or understanding of how it might interfere with other aircraft in the airspace. That’s fine if you are in a rural, unpopulated area but when you are in more urban areas with heavier flight traffic, that causes problems.
This is exactly the reason why Iris Automation was formed. Our software essentially ensures that information will be given to the pilot – so that they can take the appropriate action before a disaster happens.
Then there are intruders – drones flying intentionally without following accepted standards and protocols. Our onboard system provides an extra layer of protection and information for general aviation pilots, particularly those in helicopters who are frequently focused on special missions (fire fighting, air ambulance, police activities including search and rescue, news reporting for example) but also on the airspace around them.
It is evident that the aviation industry is your main field of focus, so what predictions do you have for the future of this sector?
This industry is aggressively pursuing the opportunities around commercial drones, autonomous flight, and enhanced general aviation. Like many industries, we have seen the increase in artificial intelligence and machine learning as a way to automate highly complex tasks and are seeing opportunities to use these to enhance operations, efficiencies, and safety. Global regulators are becoming more comfortable with what this industry can offer as testing and regulations evolve and the standards are set for what constitutes safe flight operations. Because of this we see the aviation industry morphing and growing substantially, with many more use cases and fewer barriers to entry.
Share with us, what’s next for Iris?
We’ll continue to work with our extensive ecosystem of partners, across government, general aviation, and other advanced technologies, to bring broader adoption to autonomous flight and aid those operators providing such flights to ensure they meet the high standards and regulations required.
We’ll work with many of these partners to integrate Casia into their systems for more complete solutions. We’ll continue to improve our technology, continually updating our deep learning algorithms on our simulators. For our clients, we will continue to offer them our professional services around acquiring regulatory approvals and continue our efforts with regulators around the world to accelerate uncrewed flights safely.