Engineering Safer Airspace with Ella Atkins

Ella Atkins joined Virginia Tech’s “Curious Conversations” to talk about the complexities of air travel safety. She explained the differences between Visual Flight Rules (VFR) and Instrument Flight Rules (IFR) and the challenges posed by having two sets of rules in the same airspace. Atkins also discussed the recent incidents involving aircraft collisions and the technological advancements that could enhance safety, which include updating methods of data sharing and the use of automated systems.
(music)
Travis
Air travel safety has been a surprising hot topic as we've started 2025. A series of air collisions has raised questions as to what exactly is going on in some of our most crowded air spaces and has made many people wonder what could be done to mitigate the potential for such tragedies in the future.
Thankfully, we can turn to Virginia Tech's Ella Adkins, who has an expertise in this very subject and was kind enough to share that with us. Ella is the Fred D. Durham Chair in Engineering at Virginia Tech, as well as the head of the Keventee Crofton Department of Aerospace and Ocean Engineering. Her research centers on the investigation of aerospace and aircraft autonomy and safety. And she's also a licensed pilot.
Elle explained what the rules of the sky currently are and really how there's two different sets of rules that different aircraft are abiding by at the same time. She talked about some of the challenges that presents and also shared insights as to how we might up our communications game and also add a layer of autonomy across commercial airlines that could help us mitigate dangers in the future. I also asked Elle what one change she would make today to make air travel safety better throughout the United States, as well as what her thoughts are right before she's about to step on a plane as a person who spends her life researching this topic. So you're going to want to stick around to hear those, and please don't forget to follow, rate, and or subscribe to the podcast. I'm Travis Williams, this is Virginia Tech's Curious Conversations.
(music)
Travis
I know I want to pick your brain and talk about some ways that our air spaces can be more safe. So maybe some technological changes we can make. But I think maybe a good place to start would be kind of set the landscape or maybe skyscapes is the right word for this of how air travel currently operates in our country. And specifically, I've been looking at these two. I know we have two different sets of, I guess. rules for flying, the instrument flight rules and the visual flight rules. And so I'm just curious, what are those two things and how do they interplay with one another?
Ella
The different flight rules, visual and instrument refer to the way a pilot or a flight crew of more than one prepares and executes their flight. Under visual flight rules, there's a lot of airspace that is not controlled that allow the pilot to really choose in real time the exact altitude and direction that they're going to fly. In this airspace, you don't have air traffic controllers vectoring you around other traffic that's nearby very often. Instead, you have the freedom to fly over the desert, to fly over a farm field without a lot of extra overhead associated with that. With VFR flights, you still typically will file a flight plan, especially if you're going on what's called a cross-country where you're flying more than 50 miles to your destination, but it is optional. You can choose to do that. You can choose not to do that. The reason to file a flight plan
is because that gives a level of oversight so that if something happens to you, the police will come and look for you. Whereas if you don't tell anyone you're flying, it would only be your family that would be aware that you're missing. Instrument flight rules are quite different. Before you fly, you have to file an instrument flight plan and secure approval for both the route and the altitudes that you fly at throughout your entire flight from origin to destination. That allows all of air traffic control and if another pilot is interested, they can also know where you will be. There are waypoints called 4D waypoints. The 3D is XYZ and the fourth dimension is time. So an easy way of an air traffic controller or pilot knowing where they are along a flight plan is to have kind of a spreadsheet that says, here's the waypoint I'm flying to next. Here's where I will be in space and time, when will I arrive at that waypoint? What that allows is for traffic, especially commercial air traffic that's arriving at a very busy airport to be within a minute, if not even more accurate of their arrival despite winds. You can correct for any discrepancies in expected winds on route to speed up or slow down a little bit so that your arrival time at each waypoint is as close to being exactly on as possible.
Travis
Okay, well it sounds like, and you tell me if this is wrong, but it sounds like with the instrument flight rules, there's just a whole lot more data that we have regarding these planes and where they are. then maybe the other way, there's a lot less data and there's a lot more flexibility may be provided.
Ella
So the advantage of IFR is we have situational awareness across the aviation system of what each airplane plans to do and there's a contract there. If an aircraft has agreed to that flight plan, they are responsible for following it quite accurately in space and time. With VFR, VFR aircraft can be turned down. If there's a busy area, what you have to do if you're entering controlled airspace as a VFR aircraft, you have to ask an air traffic controller, for permission to fly through that airspace. And you may say, request this altitude and flying direct, which is typical to fly through controlled airspace. The air traffic controller can either deny that request saying remain clear of this airspace, or they can approve it, but they might assign you a different altitude or route to get through it in a safe way. If you're at a really busy airport area, there's a good chance that air traffic control will deny the request of a VFR aircraft to fly in a busy airspace. But that is not necessarily something that happens all the time. A VFR pilot always has to be prepared to be denied entrance into all of the airspace types where clearance is required.
Travis
Okay, well I know that before we get on here, we were talking about some of the recent plane crashes that have happened in and around the United States, one of which was actually two aircrafts colliding. And so I'm curious, what role do these two different sets of rules play in those collisions? And what do you think we might be able to do to maybe make the airspace a little safer?
Ella
So in the Washington DC aircraft incident where a CRJ aircraft on approach to runway 33 collided with a military Black Hawk helicopter coming in the north-south direction, I think we're finding that the helicopter hit the aircraft. The aircraft was simply following what it had planned to do. In that case, the IFR aircraft had a flight plan. They had clearance from air traffic control.
Air traffic control gave them an instruction to change to a new runway, which was runway 33, and they were following the instructions that air traffic control had given them accurately. The Black Hawk helicopter instead requested visual separation. That triggers to the air traffic controller a response of approving the visual separation. And I think there's a bias towards approving visual separation and VFR flight more frequently for a military aircraft or medical helicopter or the equivalent of what is considered a critical operation than to approve that kind of route for, say, a Cessna 172 who was sightseeing. So there's a kind of a different hierarchy that even happens today. I don't think a lot of Cessnas try to mix with the jets near the Washington DC Reagan airport, but if they did, there's a good chance they would be denied. On the other hand, the military often carries VIP passengers from point to point in the DC area, and air traffic control tries to accommodate those operations whenever they can. The status quo has been that if a helicopter says, have visual separation, I want to request it, the air traffic controller will say yes. That places the full responsibility of collision avoidance on that helicopter and the pilot.
Travis
That's a really good breakdown. And that makes a lot of sense to me that we would maybe have a little bit of a hierarchy there because we do the same thing in our roads. We see sirens and we pull over.
Ella
The difference is here though, we had two different authorities, right? We had the military that says, hey, we have this helicopter, we need to fly through. Not clear about the VIP passenger thing, but that's just the way it is. The military helicopters tend to have the ability to fly through that space. But the commercial aircraft have the right of way in that space. IFR traffic like that CRJ always have the right of way to follow their flight plans that are filed and available to anyone in the system relative to VFR traffic like the helicopter. So that meant that the helicopter was responsible for avoiding the aircraft. And so the question is, if you're relying on human vision, is that okay as we move forward?
Travis
Well, I'm curious when it comes to just depending on just human vision, what are some technological advancements that we could be leaning into to not make mistakes?
Ella
We have the ability to use a combination of ADS-B, which is now a required instrument on all civilian aircraft in the United States, that delivers an output of the position of the aircraft regularly so that all other aircraft can be aware of that. And we have radar systems near the major airport that also provide position data for all of the aircraft and helicopters in the area. And air traffic controller tends to primarily use radar data. There are systems onboard the aircraft that tend to use the ADS or ADDSB data. And if every aircraft in the sky was one, detectable by radar, which in the case of human occupied aircraft near an airport tends to be the case, and B, actually broadcasting their ADSB or ADDSB data, then there would be a maximal amount of situational awareness for the people.
Now, in terms of software, we also can have that data go into software, which is extremely proficient at math. For all of the reasons we complain about software, math is a strength of writing software for a computer. And in this case, what we're really calculating is if we project forward the flight plan of the IFR aircraft and the flight path of the VFR aircraft, do those paths come close enough to meeting? that we want to have an action taken for collision avoidance purposes. Right now, we have decided to trust the people 100 % to do that in commercial and in military aviation. That means that a person on board the helicopter or a person on board the aircraft ultimately has the decision of whether to choose a maneuver or whether to just keep going exactly like they planned.
Travis
It sounds like we have the ability to communicate automatically a ton of data from our airplanes, but some of our airplanes or maybe not all of our airplanes are doing this. Is there a percentage that are using it or just none of them using it?
Ella
Also, the ADS-B data should be coming from all aircraft. For the particular helicopter incident in DC, it seems like the helicopter had their ADS-B system turned off. The military has made a statement that that is traditional because when they haul around VIPs, they don't want the helicopter's location to be known by everyone because that might increase the risk or threat to that vehicle. But there is the counter argument which now is being...considered, which is at least make the helicopter turn on its ADS-B when it's within a certain number of miles of a major airport like Reagan Airport. And then they can turn it off again once they pass that area where there are low flying jets. So I think that kind of discussion is happening and it needs to happen. The public needs to be aware that there is definitely a place like around Reagan Airport where The military traffic and their practices need to be a little bit adjusted to make sure that the passenger carrying aircraft are safe. There also are collision avoidance capabilities in software that we have not chosen to require on either civilian or military aircraft yet, but they are options. These automation aids would normally be dormant. They would do nothing. The pilot would direct exactly where the aircraft or the helicopter was flying but in a situation where a collision was imminent. Automation can turn the aircraft and have it climb or descend to avoid the other aircraft. And that is something that really is capable of being deployed as soon as the aviation community agrees that it's a good idea.
Travis
I want to ask you a question and then I want to talk about that automated technology. What does ADS-B stand for?
Ella
Automatic Dependent Surveillance Broadcast.
Travis
And so I'm curious when pilots are coming in, if I understood you correctly, they currently commercial pilots and I guess military too, let's say they're coming into land. All that information isn't just automatically being transmitted stuff like and speed and the angle they're coming in. They're I guess doing it by radio.
Ella
Yes, so they speak over a voice radio. So it has a microphone. It usually wear headphones that block out the loud sound of the airplane and you can hear what's being said. So a pilot on board the aircraft is always listening to an air traffic control frequency and monitoring what the controller is saying and what other aircraft that they can hear are saying as well on that frequency. For a busy airspace like Reagan Airport, there will be multiple frequencies. Some different for approach versus departure, the helicopters and VFR traffic may have yet another frequency that they're using. Air traffic control can hear and react to all of them. Maybe there's multiple people listening on different frequencies. Maybe you have one controller that's handling multiple frequencies at the same time. But that voice is really, it's a matter of listening to what the controller is saying and following their instructions. and as a pilot communicating clearly what your intentions are and allowing both the air traffic controller and any other pilot on frequency to hear your intentions as well.
Travis
Wow, so it sounds like it may be the difference between like having a GPS in your car to give you directions to, let's say the beach and like just trying to talk to your dad over the phone to tell you which way to go in some ways.
Ella
That's actually a pretty nice analogy, I think, because now, you know, I think the air traffic controller might be more professional and more prepared to tell the aircraft where to go than your dad would be prepared to tell you which direction to turn it. Right? So I think there's a lot of training and people are generally pretty well qualified for that job. But at the same time, we're limited to the bandwidth or speed at which people can speak. There's a lot of uncertainty, ambiguity in the English language.
Travis
I would hope so.
Ella
So sometimes people will say things that are not precise and everyone who's listening has to try to make their best interpretation of what was said. It also may be that the person doesn't hear, right? And so what happens when an air traffic controller says, take action A and the pilot doesn't hear take action A? Well, the pilot is supposed to read back what action A is and agree to do it. That doesn't always happen.
The air traffic controller, if it's safety critical, then eventually should say, hey, read back this instruction and they'll say it again. But in a very busy airspace, there's a limited number of words that can fit on that radio channel. And so it may be that they have to first tell another aircraft what to do before they get back to the aircraft that didn't actually properly read back an instruction.
On the other hand, if you have a data link where the packets of information are being sent over a secure communication channel, we can use satellites like Starlink, can use cell providers, Wi-Fi near an airport that is in a busy urban area, there would be multiple options. Those packets will be unambiguously received and interpreted. Now a lot of the pilot communities say, hey, there's all these cybersecurity issues. These are not the experts that are saying that. These are people who really just enjoy talking on the radio and see that there would be a pretty big step change in the system if you suddenly went from voice to this radio-based data link communication system. But it's a step that is worth taking first in parallel to the voice so that you can try out the data link, see how well it works. display the information to the human pilot, the human air traffic controller. And then once you get a lot of experience with how well that data link works, and it already is improving safety because maybe if a pilot just didn't hear what was said, it's being displayed in front of them. So now they are seeing the information as well as hearing it. That can already make the system more safe. And then over the long term, it's really a question that we want to address of whether we are going to eventually move to DataLink as the safety critical way of sharing information as opposed to using voice as the sole way of communicating safety critical information.
Travis
Yeah, that makes a lot of sense about how many issues that could come up between them. I think my dad would tell you that the hearing part of the equation when he's giving me instructions has almost always been the failing piece of that puzzle.
Ella
Well, are you trying to say you were maybe not perfectly behaved as a child?
Travis
I am just saying that that may be his perspective is what I'm saying. But I can't speak to it. can't speak to it.
Ella
That's okay. I hope he’s listening.
Travis
I hope so too. Speaking of some of the automated technology we have that could prevent a collision, where do we see that already being used?
Ella
Sure, so we have a system called TCAS, which is available in most all commercial airplanes today. It is an advisory system. It is not an automation aid in that what it does is it gives traffic collision alerts as advisories to the flight crew of both airplanes that are approaching each other. And it gives maybe a pull up or a push down, climb or descend type of instruction. so that both airplanes will do the opposite reaction so that they avoid each other. That purely is optional though. Some pilots tend to listen, some pilots tend to independently confirm whether there is actually a reason to divert or not, but it's still there. And that has been around for a long time, at least providing alerts to people that are flying in commercial aircraft. We also have something called Aldo GCAS, the Automatic Ground Collision Avoidance System, which has been deployed primarily on military aircraft, especially fighter airplanes. Fighter airplanes tend to fly under the radar. We use that phrase a lot when we talk about planes and flying under the radar. Well, what that means is that you fly close enough to the terrain or ground that radar systems will not track where you're going that poses a lot of risk to the aircraft and any occupants of that aircraft because if you make even a little mistake flying really fast, you can hit the terrain and that's the end. So this system was really motivated by the desire to reduce the number of controlled flight into terrain or CFIT accidents by providing automation that pulls the aircraft up to avoid terrain. even if the pilot does not take action.
Travis
Wow. So it sounds like that it works against terrain. Could that also be employed to work against, say, a mountain or some other obstruction that a plane might be headed towards?
Ella
Yeah, so a mountain is part of the terrain. It could also work against buildings and so on, right? So I think there's a combination of maps and radar signals that would be able to provide information that would guide it. There's never a good reason to fly, at least in civilian terms, a perfectly good aircraft into a mountain, a building, a power line, or anything else you always want to get to your destination and have your passengers exit safely. And having the passengers exit safely is thing one. Getting to your destination is thing two. So if you need to land at an alternate airport, passengers may grumble about that, but they grumble safely.
Travis
That is true, as long as they're still able to grumble, think that's a good place.
Well, if there was maybe one thing you could do today to make our air spaces safer, what would it be?
Ella
My first thing is that data link. We have to be able to use the same technology that we trust every day in devices like this, cell phone that I'm holding up to communicate. We trust it to navigate. We trust it to in so many different ways. We feel safe sending our kid out to some event because they have their cell phone and they can contact us if they get in trouble. Yet we are afraid in aviation to do the equivalent.
It's time to stop being afraid of data link and let it be what the year 2025 suggests it can be. There is no reason not to share safety critical data with every airplane and every air traffic control personnel and every computer system in air traffic control and in the cockpit of every aircraft so that there is situational awareness shared by the people and the computers which collectively will keep the system safer.
Travis
Well, knowing all the things that you know and you study in this area, does that change how you approach air travel?
Ella
I will get on a CRJ in Roanoke the next time I fly most likely because I tend to fly on an airline that doesn't go a lot of different places from Roanoke. And so I will feel just as safe, maybe safer than I used to because there was the Toronto accident that appears to have been really showing the masterful engineering of the CRJ aircraft, right? Because nobody died, even though it flipped upside down. So I consider there to be some accidents like that that really show and highlight the quality of aerospace engineering. And as the department head of aerospace and ocean engineering, I feel very proud of all of the engineers out there who have contributed to making airplanes safe by design. So that's where I'm gonna be in my mind the next time I fly. And I will not hesitate to go there. In fact, one of the interviews I did right after the DC crash, I was driving to Charlotte to get on an airplane. I pulled off at a rest stop and talked to one of the media outlets And they said, hey, it looks like you're in your car. And I said, yes, I'm on my way to the airport and I'm going to sleep well on this flight because I feel just as safe today as I would have felt last week.
(music)
Travis
And thanks to Elma for sharing her insights related to air travel safety. If you or someone you know would make for a great curious conversation, email me at traviskw at vt.edu. I'm Travis Williams and this has been Virginia Tech's Curious Conversations.
(music)
About Atkins
Ella Atkins is head of the Kevin T. Crofton Department of Aerospace and Ocean Engineering and Fred D. Durham Chair in Engineering at Virginia Tech. Her research centers on the investigation of airspace and aircraft autonomy and safety, and she is also a licensed private pilot.