Skydio, a leading US-based drone manufacturer, proposes that autonomy is one of the most important developments for the drone industry today. In their latest webinar, “Accelerating Your Part 107 Business with Skydio,” they explained why.
“Everyone agrees that drones are ten times more effective and 100 times cheaper and safer than heavy machinery,” pointed out Guillaume Delepine, Product Marketing Manager at Skydio. “If you compare drones against the cost, time, and the risk to safety and life of traditional methods it seems like a no brainer that we would be using drones all the time, but that hasn’t totally happened yet. Why? For starters, if you look at a lot of the most developed programs, even at scale, about 80% of your budget will end up consumed by pilot training and salary.”
Because manually piloted drones are so easily crashed, they require a hefty investment in qualified pilots and visual observers, as well as pilot training to try to lower the risk of pilot error that would lead to loss of equipment, property damage, or worse. This cost is prohibitive to scaling many drone programs to their full potential, which is one of the reasons why heavy equipment like helicopters and snooper trucks are still being used despite their cost and safety risks.
“The cost of the pilot in the field and a visual observer is substantial,” explained Trevor Ragno, Chief Real Estate Officer of Aeronyde Corporation, whose company runs a multi-pilot program that has started to ramp up their use of the Skydio 2 system. “We found it took up to 82% of our operating budget. This is why we saw autonomous systems as an essential component to a full enterprise solution.”
Seeing this large operational cost as a blocker to industry growth, Skydio is working to bring about “the age of AI-driven autonomy,” where drones are no longer powered by manual operations but are defined by software and AI with native obstacle avoidance. Skydio argues that fully automated workflows and integrated solutions, where drones can make split second decisions in the air to keep operations safe, will reduce operating costs, thereby enabling the enterprise.
To this end, Skydio has developed a full suite of enterprise solutions designed to usher in an autonomous workflow for drone operations.
Onboard every Skydio 2 aircraft are six 4K cameras with 200-degree fisheye lenses that collect 45 megapixels of visual sensing, making them the eyes of the Skydio Autonomy™ engine. The brains of the Skydio consists of a NVIDIA Tegra TX2 processing chip, which is capable of doing 1.3 trillion operations a second.
“The NVIDIA TX2 board is essentially a supercomputer on wings,” said Ragno. “To put that into perspective, if you were to spend a million dollars for the last 2,000 years, you wouldn’t be halfway to 1.3 trillion today.”
These combined features enable the Skydio 2 platform to see, understand, predict, and act on the world using artificial intelligence. It creates a real-time, three-dimensional model of the world as it flies through it with 360-degree obstacle recognition and avoidance, and motion prediction.
“By recognizing objects and their context, it fills in the gaps that a sensor might not pick up,” explained Delepine. “It can predict your motion for accurate path planning or predict the motion of other subjects it is following, gathering data that pilots can use to get the job done, and then it is automating entire workflows—all of these factors are absolutely critical, without it, it can’t be called an autonomous drone.”
On top of the Skydio Autonomy™ core engine, Skydio offers an add-on solution for AI pilot assistance they call Skydio Autonomy™ Enterprise Foundation, which includes super zoom, vertical view, and Close Proximity Obstacle Avoidance, among other features, all of which help the pilot increase the level of situational awareness, as well as help them fly into tighter spaces where a manual pilot or human couldn’t, shouldn’t, or don’t dare to go. Skydio’s AI powered flight software according to Delepine can “turn anyone into an expert pilot.”
Skydio is also working on a 3D digital scanning automation solution—Skydio 3D Scan, which will provide a fully adaptive scanning solution for photogrammetry capture. This solution, along with Skydio House Scan, which Skydio delivers in partnership with EagleView, is what Delepine refers to as fully automated workflows, where the drone functions as an edge IoT device that can completely automate the data gathering process, dramatically improving efficiency for Part 107 pilots working in inspection-related projects.
Skydio’s solutions are already making a difference for end users in multiple verticals. Being able to fly a drone in a single continuous mission while being able to fly at low altitude without the fear of the drone crashing, has made a huge difference for Sam Delong, CEO of Accurate Drone Solutions, a drone service provider that offers construction site mapping services.
“It’s unreal what this drone is going to do for construction and the industry in the future,” said Delong “Less flight time is a huge one. Being able to create a pattern and have the drone execute it without having to pause the mission to start at different waypoints is huge. Being able to maintain a consistent altitude, while also having a drone that actively senses the environment around it in 360 degrees is a complete game changer for enabling safer operations. And the quality of data it generates is bar none.”
The ability to capture high quality data in one flight, also makes a significant difference on costs and scalability. When a drone doesn’t collect the data well the first time, pilots have to recapture that data. Depending on when the data was processed, this may involve multiple trips to the data collection site. Factoring errors into your projections for work can have a major impact on what you agree to do and how far you can scale. Therefore, having confidence you will capture the data right the first time can enable businesses to take on more work.
“The definition of risk is how much you’re going to deviate from expectation,” explained Ragno. “This isn’t just about the risk of crashing a drone, which is also very real when you don’t have obstacle avoidance capabilities like the Skydio, it is also about the risk of misquoting someone, especially if you have a guaranteed maximum price. Depending on how far you have to travel to the site, or how long it takes you to gather that data, you really want to make sure that you get the job done right the first time and Skydio helps us do that. In our case, I don’t think we’ve ever had to go and refly a job with the Skydio.”
Beyond safety and legal ramifications, mitigating risk can be quantified by how much money you make in current jobs and whether you will secure more jobs in the future—two major things to consider when starting a Part 107 business with a drone. Drones, like Skydio, that have reliable autonomous capabilities and can produce the highest possible deliverables is really what is going to enable the enterprise of the future to grow—replacing expensive heavy machinery and dangerous operations.