A drone program at the New Mexico State University (NMSU) test site caught my attention. They are doing safety work: Find small drones, track them, and reduce collision risk.
The problem zone is surface to 400 feet AGL. Helicopters work there. Manned aircraft pass through on climb and descent. Now add the drone stack and more low-altitude traffic coming, including eVTOL operations. Same layer. More mixing.
See and avoid does not hold up well here. Drones are small and they blend into the ground. In the drone zone you are also watching terrain, wires, landing spots, and other traffic. It can be harder than higher airspace because everything is closer and the targets are smaller.
ATC may not see the drone at all. Most drones are not on ADS-B, so you usually do not get a target on your screen. That is why I keep pushing DFR / DFR-B.
Digital Flight Rules (DFR) means digital separation. Not “look harder.” Not “ATC runs every drone.” It means a rule set a system can follow, with early conflict moves built in. Drones should also be in a network where they can talk to each other, so they can avoid each other, not just hope the human sees them.
Rules come first. Code follows. DFR needs clear answers: who yields, when they yield, what spacing is required, and what the vehicle does when something fails, like a lost link or bad navigation.
That is where DFR-B fits. DFR-B is DFR for the band, surface to 400 feet AGL. This is the urgent layer. Risks are close to people and property.
DFR-B also cannot be a separate world. Helicopters live there. Manned aircraft pass through it. Some unmanned aircraft go higher and stay up longer. The rules have to connect upward into the larger DFR idea or the handoff becomes the weak point.
New Mexico has a good test range for this. NMSU’s UAS Flight Test Site operates under a Certificate of Authorization (COA) that permits UAS flights in over 15,000 square miles in southwestern New Mexico. That airspace runs from the surface to 17,999 feet MSL, with low traffic over mostly undeveloped, government-owned land.
NMSU shows the drone-zone issue is real enough to build detection and tracking tools. The NMSU UAS test site (UASTS) shows unmanned testing is already happening. What we still lack is a clear set of separation rules that works across the whole mix. That is what DFR (regs) and DFR-B (system logic) are aiming for.
Note: The acronym DFR-B is my own. It is purely for the purpose of separating DFR (Rules) from Digital logic (DFR-B).
Before we know it, another box will be mandatory to be installed. There may be another deadline that will force people to comply or simply cease exisiting/ operating in the drones preferred airspace.
In my opinion, anything autonomously operated needs to have the capability to see (or not see) and AVOID anything with a pulse and heartbeat flying under its own power or occupying any contraption built for the purpose of flying. Yes, I mean birds, too.
Someone smart in the insurance industy should start some extra smart predictive modeling for the total number and frequency of potential claims and also the cost/ severity for the insurance world.
Very cogent points, Jason. Any solution that involves ADS-B, will require some aircraft (both of mine) to buy and install additional hardware to fly in airspace that is now open to me. Throw in the avionics cost and that’s a serious burden to accommodate a threat that was never there before.
This should be a “cost of entry” to those who are encroaching, not us.
Best answer I’ve seen so far is DFZ: Drone Free Zone from the surface to 500 ft AGL (400+slop) on all runway approach and departure paths. Also, all Ag areas are Restricted for drones… find and contact the land owner for permission to enter.
As there is already a working system in place, theoretically ADS (In/Out) should be the first place to look. Strip the physical containers and fancy screen for a moment, ADS-B OUT is nothing more then a transmitter and some software. RC planes today carry this capability so it is not so much an issuer of a lot of added mass. ADS-B IN, receiver can be built in and again, the working parts are software.
Yet looking on-line, even basic units are minimum of 900$ and up. Is this because is aviation so of course is expensive, because of supply/demand such that there are not enough planes to drop costs on volume?
I looked up ADS-B receivers for home, lowest cost was 35$ with many still under $100 so clearly a receiver could be added with code that states, if anything gets near you, avoid it. if the cost can be so low for receivers is this then regulation issue or some BS privacy issue (I don’t want the ‘Ment tracking my stuff).
If IN is that cheap, OUT should not be any more so but (I guess) regulations relation to transmitting signals, but here again, Model rocketeers have been putting in real time telemetry reporting in rockets for years. An Arduino with a basic telemetry transmitter is less then 50$ so it cannot be a hardware cost.
As an RC enthusiast I see a difference between standard RC planes (aircraft) used as a hobby compared to drones which more and more are used commercially and non-commercially, but for Social Media postings. I would oppose requiring hobby RC planes and helicopters from being required to carry some type of ADS-B hardware, but any commercially used drone should, any drone over a certain mass or a drone that can exceed distance from transmitter of 2640 ft and 400ft vertical be required to not go further or carry ADS-B.
I agree we need Rules for drones, starting with what defines a drone from other RC controlled objects then look how technology can help in separation and usage. There will always be bad players trying to break the rules, but you are correct that we need them now, not later. (kind of like AI as well)
Good comments all. As to rules, all drones should be burdened (get out of the way). ADS-B IN with simple conflict avoidance (3Tau) adds little to cost of serious machines. Toys and recreational RCVs must operate under local see and avoid rules.
Folks, we don’t need to reinvent this wheel.
Isaac Asimov did it for us 84 years ago in his short story “Runaround”, later incorporated into the collection “I, Robot” with which you may be more familiar. (And please don’t confuse his story with that execrable Will Smith movie.)
This brilliant and prescient little story laid out Asimov’s “Three Laws of Robotics” which were designed to safely incorporate intelligent machines into human society. To wit, and with only one word substitution:
These three simple rules, to a limited degree, are already being incorporated into the programming of industrial robots, insofar as possible. Where not, humans are blocked from approaching. The problem with that solution is that it places the onus for human safety on the humans, not the threat itself. Therefore industrial robots are limited to very tightly controlled environments with limited human contact. That won’t be the case when the skies are full of effectively invisible service drones, which I dubbed “autonomous aerial mines” years ago.
In a prudent society, no autonomous mechanism would be allowed in human spaces without such safeguards. But we can’t seem to ensure that the human-carrying robotic cars on our city streets are able to close their own damn doors. And don’t get me started on “Roomba Fails”.
Put another way, who is responsible for a bullet wound? It’s pretty hard to claim that it’s the victim, and impossible to blame the bullet itself or even the gun. That leaves only the person who pulled the trigger. Yet here we are, all set to “blame the victim” for not properly equipping his flying machine with expensive defensive paraphernalia.
The FAA’s lazy reductio ad absurdum approach would solve the gun violence problem by requiring everyone walking outside to wear medieval chain mail and armor.
The solution does not lie with the potential victims. The onus is squarely on those who want to unleash this new threat upon them.
-Chip-
I agree, Aviatrexx. Pilots already using the system should not have to suit up with more gear, more electronics, and more cost just because a new class of machine is being pushed into the same airspace.
Your premise is valid BUT … what / how does DFR-B work and interface to the existing system, Raf?
Not in the existing system; the BNACTS+DFR-B system. See comment after your “crows into a prison” story. 😂
Where is AOPA on this, or will they be touting the next $5,000 box we weekend flyers have to install as the greatest thing since pockets on a shirt, like they did the last time around? One wonders if Garmin and Honeywell execs sit on AOPA’s board.
Here’s a ‘new’ use:
“An unusual smuggling case has surfaced in Louisiana, where two women have been arrested for allegedly flying drones disguised as crows into a prison. Texas natives ‘X’ and ‘Y’, have been accused of sending phones, tobacco, and drugs, including methamphetamine, into the Grant Parish Detention Center.
The Grant Parish Sheriff’s Office (GPSO) has shared an image of the decoy crow, showing a plastic toy covered with black tape and fitted with eyes and a black beak. The dummy bird was connected to a drone, and a white bag was pasted on its back.”
“Larry S: Your premise is valid BUT … what / how does DFR-B work and interface to the existing system, Raf?”
Larry, DFR-B is a low-altitude separation concept for the drone band, but it is not meant to be fenced off. It applies across the entire National Airspace System.
It should connect through the NAS network side, not by adding more cost and equipment to the pilot already flying there. One way to do that would be for BNATCS to include the traffic logic side: shared traffic picture, digital constraints, and rule distribution.
The drone manufacturers and operators would then carry the onboard side: intent sharing, detect-and-avoid, yield logic, and fail-safe behavior.
The idea is that DFR-B could run independently for separation, while still staying networked for situational awareness. In other words, the unmanned side should not have to wait for ATC to solve every conflict, but it still has to share enough information to fit safely into the larger system with focus on the drone band..
EASA, the UK, and the FAA are all working this problem in their own ways, but there still is no uniform worldwide system. That is even more reason for the FAA not to fall behind while the drone population keeps growing.
If unmanned traffic is coming into the system in force, then the unmanned side should bring the compliance logic with it. Existing pilots should be protected, not retrofitted.
That is the direction: logic in the system, not more boxes in the cockpit.
I agree with the reasoning for the unmanned side to bring the compliance. I still don’t see how the interface with the NAS will work, tho. If a drone was squitting ‘electronic conspicuity,’ wouldn’t they clog up the ADS-B frequencies? And what about airplanes who don’t — as of now — have to transmit ADS-B out? How would the drone know there’s an airplane nearby? I foresee the order to transmit ADS-B out for everyone in all airspace. And … when a drone sniffs an airplane in some predetermined box, they’d have to go down very low OR stay out of certain airspace as
I see a can of worms forming
Larry, I have been posting on this subject in AvBrief to bring the safety problem and possible fixes into clearer view. This is not a small issue. We are already dealing with millions of drones, from small consumer units to larger unmanned aircraft and eVTOLs, and that number is still growing.
My point with DFR-B is simple.
If unmanned traffic is the new load entering the system, then the unmanned side should carry the responsibility, logic, and software needed to avoid each other and avoid manned aircraft too. I am not looking to put one more burden on the pilot in the cockpit.
I understand the squitter problem too. The Albuquerque Balloon Fiesta showed that even a few hundred balloons in one area could turn ADS-B into an operational problem that needed waivers and workarounds.
So no, I do not see ADS-B as the answer to the drone volume coming. If a few hundred balloons can strain the system, then millions of drones squittering down low is not a clean solution. It is a scaling problem.
NASA, EASA, and RTCA are all working parts of this problem, but NASA comes closest to the DFR-B idea. EASA still leans more toward managed services and some cooperation from manned traffic. RTCA is writing the hardware and software rulebook, but not really taking on the bigger burden question. And the FAA still looks behind the engineering curve. Plainly, the side growing into the airspace should be built to operate safely there.
What I would really like to see is a serious engineering panel with uAvionix, Garmin, NASA, RTCA, and the FAA in the same room working this problem, putting the standards, certification issues, and cockpit realities on the table, and explaining how unmanned aircraft are supposed to separate from each other and from manned traffic.
Got it. I didn’t know about the balloon overload in ABQ but it doesn’t surprise me … hence my former comment.
Cockpit overload with all the information being presented on the ‘TV sets’ in modern airplanes is real … dealing w a hoard of drones below 400′ would only make it worse. I see a time when airplanes will consider dropping below 400′ altitude as descending into a ‘soupy mess’ and playing Russian roulette.
I foresee the first step as requiring all aircraft to have ADS-B out in all airspace. (I don’t agree with this; it’s just added $$ for those who don’t need it). Electronic conspicuity might help the Cub drivers? IF a drone could ‘hear’ and avoid an airplane, that’d be a start. Requiring drones to carry a ‘modified’ ADS-B out might work IF some special coding were applied to their transmitted low power signal ID’ing “drone here” such that airplanes would ignore it until the aircraft’s altitude approaches that 400′ drone ceiling. Slight modification of ‘in’ boxes to ignore these signals until they became relevant would be fairly easy, I’d think?
There are a lot of permutations on the ideas BUT … I agree with JaBa and Aviatrexx and you that the brunt of any cost or accommodation should fall upon the drone operators. TBSS’ idea of Drone Free Zones is also another idea in the vicinity of airports. Required geofencing drone navigation (already done) would help. I could see a system where the FAA might transmit localized signals to all drones modifying geofencing as needs pop up … e.g., a sports event or something special.
The DCA crash is an excellent (sad) example of where all the technology existed, the aircraft involved had the equipment to see and avoid (plus their Mark I eyeballs) but one wasn’t using it (under a BS exception) and therefore was the primary reason for the midair. There were four pilots plus (overworked) controllers that shoulda / coulda avoided that crash … but didn’t. So I don’t have any great hope that technology is the ONLY answer here.
Frankly, I don’t see where it’s necessary to have one’s BigMac or detergent delivered by drone. Slowing down superfluous uses would help. Get offa you buttski and go get your stuff. Maybe drone delivery to a central location for local pickup is another idea? Drones flying down predetermined ‘airways’ to drone “ports” might be another idea, then.
What a mess.
Many of you remember the Moller Skycar. Well, around 1998 to 2000 an inventor took the Skycar concept to the next step and created a kit that changed your car’s stock wheels to generator/wheels. Then bolted four electric motor fans to the roof of most any production car. Although it was really cool to watch it do the wobble tether hover, it grabbed the FAA’s attention.
The FAA posed the question to the Alaska University near me; “What if everyone installs these by the millions, what and how do we control air traffic?
I sat in one of many work groups and brainstormed ideas. The one that kept coming back to the discussion was the “Traffic Circles” around the city and within the city at different levels. The Air Traffic Control (ATC) would release you from the circle and the ATC would fit you into the left rotation. The many lower altitude circles are slower speeds and smaller radius and with each altitude the speed and radius would increase. About 5 to 7 minutes to go full circle. Without getting into all the details it was possible to do with just the technology of the day. Before GPS was considered as an option. No EFIS or Moving Maps either.
The bolt-on kit was fairly simple and stock gas engines produce more then enough electricity to hover. Computers weren’t up to the challenge at the time but, things are changing.
Good point, Klaus. That Alaska work group was dealing with the same basic problem and was ahead of its time. Those traffic circles by altitude were a practical way to manage flow, spacing, speed, and entry into the system.
DFR-B would try to do the same thing with digital routing, automated conflict avoidance, altitude bands, speed classes, and entry-release logic. Different time, but a much bigger traffic load now.
Is it naive to point out that the problem of safely integrating autonomous vehicles into an already crowded space has not been satisfactorily solved in 2-space? Why the hell is any of this even legal?
Oh yeah. There’s money to be made.