A debate unsettled
My roommate and I are both aviation nerds, and often, late at night, we will get into a debate about the future of the industry. One area of interest that we often debate is drones, autonomous aircraft, and EVTOL (electric vertical takeoff and landing vehicles). The debate often starts with his prognostication about autonomous vehicles taking over the skies, replacing cars and trucks with a flying equivalent. I must admit, it’s an attractive view of the future. Goodbye traffic, hello quick flights across the city. No more laying on the horn incessantly despite the futility of two six lane highways trying to merge in the heart of Atlanta during the middle of rush hour. Next, we’ll debate the merits and potential of each of the aforementioned vehicles. Will it be an existing helicopter or aircraft that’s retrofitted with new technology or will it be large tilt-rotor quadcopters? The debate rages on. Lastly, without fail, we will end up at our final debate – how quickly they will arrive. This is where me and my roommate have our biggest differences. With investment from large tech and aerospace companies, my roommate is quite bullish on speed to market. I on the other hand, take a more cautious approach. I worry about how the FAA, the federal agency responsible for approving all aircraft and then ensuring they are safely routed from location to location, will be able to handle the spike in new aircraft. I question how their systems will, which are built on radar and human air traffic controllers coordinate the safety of all these aircraft. Luckily, there are companies looking to solve this issue. However, before we jump into the solutions, let’s put into perspective the cataclysmic shift that potentially awaits our air traffic system.
The rise of autonomous aircraft
Growing up watching the Jetson’s, flying cars were always the image of the future. And for the majority of my life, the promise of mass flying transportation was always promised as “10 years away”. Well, after multiple decades come and gone, it seems that we have finally hit the point where the countdown to mass flying transportation has arrive. Logistics companies like UPS and Amazon are racing to be the first to redefine last mile delivery through the use of drones. Ride sharing companies are racing to take the experience to the skies, with companies like Blade, Voom, and Uber leading the way. Not to mention the countless other companies that are developing drones, autonomous aircraft, and other flying vehicles that look to reinvent the future.
To put the size of this growing market into perspective, Research and Markets predicted in 2017, that the EVTOL market would grow by nearly 5x from 40M in 2017 to nearly 200M in 2026. Even the FAA, which had modestly predicted 452,000 small unmanned aircraft systems (sUAS) in 2023, recently revised their prediction, estimating that we would reach that same mark in late 2019 – early 2020, nearly 3 three years ahead of time. Compare that number to the approximately 8,000 commercial aircraft in the US commercial fleet or 190,000 in the civil aviation fleet (think smaller propeller planes) and you quickly can see just how big a shift this truly is.
Why this presents an issue for the FAA
The FAA is responsible for a number of institutions that look after the safety of our aircraft and airspace. To begin, they are responsible for the oversight of aircraft certification, the process that sees a test aircraft be allowed to fly in the sky on a regular basis. The second institution they are in charge of is air traffic control, the system that provide aircraft authorization to takeoff, fly in a certain direction, and land. It is my belief that both of these institutions will be severely taxed by the introduction of autonomous vehicles.
Let’s start with aircraft certification. While this is typically a process that is out of sight from most everyday consumers, the difficulty of safely testing and subsequently certifying an aircraft has come under intense scrutiny with the recent crashes of the 737 Max. Without getting into the nitty gritty of the certification process and what may have gone wrong with Boeing, let me use it as an example of why aircraft certification is so difficult. A typical aircraft today may have millions if not billions of lines of code, hundreds of thousands of parts, and a supply chain that extends globally. In order for an aircraft to fly successfully all this code, parts, and supply chain must work in unison to meet exact standards. Then, engineers must predict which systems might fail, which ones are critical to flight, and ensure there is redundancy or a margin of safety. The whole process is a delicate balance. The good news, aircraft manufacturers have quite a deal of tailwind in getting this right. Gone are the days of pushing an aircraft off a cliff and hoping it flies. Instead, they have a century of sophisticated research, proven aircraft designs, and advanced testing techniques that ensure they know as much as they can about the airplane before it even flies. As a result, today, flying commercially is one of the safest modes of transportation available.
So let’s play this out in the world of UAVs and autonomous aircraft. Today there are numerous manufacturers of these UAVs, some with the backing of traditional aircraft manufacturers and some independent. While some use similar designs, other manufacturers are coming out with novel configurations. Some might be using open-source software, while others may be using newly developed software. Some might be using tested and proven hardware, while others may be using less proven hardware. Yet all of these aircraft will be ultimately be airborne, meaning that they fall under the responsibility of the FAA. While most of these aircraft will be small and pose minimal damage if they crash, the risk of crashes for these aircraft cannot be completely ignored. Imagine driving at 60 mph down the highway and having a five pound projectile traveling at its terminal velocity crash into your wind shield. Or perhaps the code on the aircraft still has bugs, and the aircraft takes off and hits a pedestrian. Are these situations unlikely, yes, but can they be totally ignored? No. On the other hand, thinking about larger UAVs and autonomous aircraft, which fall under more intense FAA scrutiny, the questions become both about ensuring the safety of the aircraft and also about the ability of the FAA to process the certifications in a timely manner. Can they FAA’s systems and processes, which are built to test a limited number of new aircraft, be effective when its dealing with a swarm of manufacturers, each with an entirely new aircraft? Will the FAA’s certification be the bottleneck on innovation in aviation?
Air Traffic Control
The other key institution that the FAA is responsible for is air traffic control. Air traffic control is responsible for giving out directions to aircraft and ensuring their safety both on the ground and in the sky. They give everything from takeoff instruction to landing instruction and everything in between. It is their ability to quickly process air traffic that allows our airports and airlines to operate as efficiently and safely as they do today. Yet, for all the advancements that have been made, the final delivery of instructions is in the hands of a human, who can ultimately only process so many aircraft an hour. While it’s possible to hire more air traffic controllers, restrictions around how close aircraft can fly ultimately limit the number of aircraft that can enter a given airspace.
As drones become more prevalent, airspace will likely become more and more crowded. On one hand, many of these UAVs are small, hobby aircraft that don’t fly above 100 meters and stay within eyesight of their operators. For the time being, let’s assume these aircraft pose a minimum risk. On the other hand, there are a number of aircraft that fly well outside these parameters, potentially entering the airspace of other aircraft. These aircraft will need to be managed carefully by air traffic control and will ultimately need to be integrated into the systems (such as radar) and process that air traffic control used today. Given how stressed the system is today, adding all of these aircraft presents a massive problem for the future safety of air traffic control.
Startups looking to solve the issue
Ultimately, I suspect that speed of adoption of UAVs, autonomous and vehicles will be just as tied to how quickly innovation occurs in aircraft certification and air traffic control, as it is tied to innovation around autonomous systems and aircraft design.
Startups that are looking to address this space will have to think creatively about ways to leverage technology to reduce the workload for humans through automation, while also ensuring that the safety levels that have led to global trust of aircraft is maintained. One startup of interest that is looking to tackle this problem is SkyGrid.
Backed by Boeing and SparkCognition, SkyGrid is looking to solve air traffic control issues through the use of AI. Using a distributed system, where each aircraft is responsible for monitoring the surrounding airspace, SkyGrid is looking to automate flight routing, conflict management, and inflight monitoring. Instead of simply trying to circumvent today’s air traffic control, SkyGrid is also actively working to integrate with regulatory agencies, creating in essence, a private extension to the FAA’s public arm.
Ultimately, the success of SkyGrid will be dependent not only on their own innovation, but on the willingness of government entities and the flying public to trust companies like SkyGrid. While pushing boundaries of aviation is important to the industries growth, innovation must be accomplished in a safe manner that respects the hazards of flying and the credibility of the industry. While there are many unknowns left to be answered, the fact that new technology is pushing our existing institutions like the FAA to innovate, should be seen as a positive sign for the industry and a call for new and existing payers to implement technology in new, exciting ways.