UAV swarm, anticollision - faa - detect-and-avoid

Since UAV activity is largely constrained by the see-and-avoid issues;

I though it would be appropriate to propose a solution which is compatible for every form of flight from gnat-weight to jumbojets. I've worked this out is a simulator - but here is the jist:

This requires only a single frequency and very short blips.

1. Every plane determines it's location and vector by GPS.
2. Then it transmits this information using one short blip every second or so.
a. The blip timing is based on current GPS location, so that the blips cannot interfere, and their location in Timespace tells us where they are coming from. Since GPS provides a very accurate, shared time base. So let t in microseconds = latitude\1 + longitude\1 * 100 + Alt in Km *1000 .
b. Odd seconds communicate vector (ie destination), Even seconds communicate Location.

3. There is a third variable required - which I call "Density". Each plane calculates the perceived Density of their airspace and sends that as well. Your density is your passenger count divided by the distance between you and other planes multiplied by their perceived density.

4. For manned aircraft nothing further is required - this beacon signal will ward off all unmanned vehicles.

5. For Autopilot or UAV, it is possible to chart a new safe vector by trying random vector changes and calculating their future density based on the vectors and densities of nearby airplanes. The vector which optimizes progress toward destination with lowest density is the vector chosen.

simple calculation requirements (ie Arduino), cheap transmitter/receiver requirements, and low power requirements. cheap unit cost, and no ground control.

Such a system could be demonstrated by coordinating a swarm of Ardu-Planes.


Views: 2268

Reply to This

Replies to This Discussion

I hear you on the reinvent - but of course, DIYDrones is all about reinventing the autopilot in a form which is open-source, and cost minimal. I started this thread because it seems that a means of coordinating drones, and avoiding collision - which is consistent with these goals may be a useful contribution to that end.

It seems that FLARE and the other systems mentioned are either expensive and proprietary - or they do not provide for full ATC functions, such as scheduling landings, establishing directional corridors, and regional-scale congestion avoidance.

The sim unfortunately is in an old harddrive - but I need to get it out and dust it off.
This system could be added to the Ard-UAV for approximately $30 in radio parts - that's more than an order of magnitude less than the $500 low-end for some other systems I see - and almost two orders less than the $1700 price point of these other links.

I see from that link Howard I was mistaken about the operations of the TCAS. I was assuming the TCAS had some sort of limited power transponder and I see now it is, as you stated a passive system.

So it only meets half of the See and be Seen goal :-)

None the less a 1030 MHZ reciever and decoder would be a useful addition to any autopilot system.
For those who share an undying worship of the homo-sapien as flight coordinator extraordinaire - MSNBC is breaking this story:

BREAKING NEWS: At least 7 killed when medical helicopters collide over Flagstaff, Ariz.

These, and similar news chopper collisions, are exactly the kind of collisions, a near-field peer-to-peer anti-avoidance can resolve. Indeed - the incident-oriented flights are the most likely to collide since they include a motivation to share airspace, and are least likely to be ATC coordinated.

Give it up for the FAA, for bureaucratic rather than mericratic government, and for every one of the GA pilots who have stood in the way of an improved system. I suggest that we carve the name of these "see no progress" Pilots into these seven caskets.

This collision could have been avoided for $50 in parts from Digikey. I guess that is too much to spend on four pilots and 3 medics.

There seems to be a mistaken impression in this thread that ADS-B requires human ground controllers in the loop for anti-collision purposes, in fact, it does not. ADS-B transmitters broadcast the aircraft position and velocity vector once per second, a nearby ADS-B receiver can hear these transmissions and autonomously identify a possible collision threat, just as you suggested, even if there are no ground stations in the area. The ground station network receives ADS-B transmissions, merges them with radar-derived position data from transponder equipped aircraft, passes it off to ATC for use in their controller functions, adds additional weather and airspace information, and retransmits the merged traffic data on the two separate frequencies used by low altitude/speed aircraft, and high altitude/speed aircraft (why there are two frequencies is a long story, but in short, the second frequency was added to reduce power consumption, complexity, and cost for GA aircraft).

I'm afraid $50 worth of parts simply will not do the job. Considering the possible closing rates between aircraft of various types and pilot reaction times, a minimal range of 5 to 10 miles is needed for a viable system. The highest power XBee modules, for example, have a specified maximum range of 1 mile. Given that most aircraft installations will be less than optimal from all possible aspects, even 1 mile would be wishful thinking. You'd be looking at something more like the XTend modules, which you'll note, are considerably more expensive.

Finally, $50 worth of parts generally needs to be multiplied by 5 to 8 times just to get a finished product, without including amortization of R&D costs or even a minimal profit margin. Parts usually don't assemble themselves into a product for free.

I don't know of any GA (or even airline) pilots "who have stood in the way of an improved system". The FAA is a bureaucratic nightmare, but in reality, when you are talking about engineering a system which, if it fails, could results in the downing of an Airbus 380 full of passengers, a bit of bureaucracy can be a good thing.

I don't want to be discouraging, building a system like this for swarming UAVs would be a lot of fun. But, if you want to integrate with the rest of the air traffic control system (which is subject to FAA, FCC, and, in some cases, ICAO regulation), it will be extremely difficult (if not impossible) to get a per unit price of less than $1000US.
I live in Az and saw the news of the helicopter crash last night on our local evening news.


I was actually watching the live news coverage of the police car chase last July when the 2 news choppers collided and our TV's went blank as they lost their feed as their helicopter fell to the ground.

Both the incident last night and last July were related to helicopter pilots knowingly trying to share the same space (or nearly the same). I don't think an automated system of any kind would have helped.

In last nights case it appears they were both jockey for the same airspace, one taking off and one landing. In Last Julys case the pilots were on the radio to each other and fully aware of each other and the assumption is they were ignoring safe separation to get a good TV shot. Perhaps an automated system would never have let them get that close to begin with.

Don't get me wrong I am a huge fan of a peer to peer CAS (Collision Avoidance System) and it would greatly increase the usability and safety of the NAS (National Air Space) but there is no way to make a pilot obey unless you take the pilot out of the loop and that brings up a whole bunch of other issues.

bGatti I think I read some pretty cool stuff on Swarm, Co-ordination, and Collaborative Control of Unmanned Vehicles by this Gal.
It was a while back but I think they used Xbee networks.
Marc that was me that made the comment on ATC in the ADS-B loop. Again you have provided me with the correct info. It was a very long time ago that I did that research and am not fully up to date. I am not a Pilot but my father, most of my friends, and co-workers are, and I will be when I can find the time :-)
I think ".... stood in the way of an improved system" might be too strong. I know several older pilots who are extremely resistant to change and what they consider interference with GA by the FAA and "Big Aviation" interests.
I think the need for an ADS-B or a Peer to Peer (Like bGatti's) type system is urgent and I don't think autonomous UAV's will be allowed in ANY Airspace until one is in place and even then I would be willing to bet FAA requires a live controller to be responsible for that aircraft sort of like a remote PIC (Pilot In Charge) to handle things that the drone can't. Perhaps one controller responsible for maybe hundreds of drones over thousands of miles.
I am a rank amateur, and I have figured out how $50 in parts could provide an effective Peer-to-peer collision avoidance. If I can do that in a few days; then the FAA is stalling, or criminally incompetent. If the FAA had moved to computers-in-the-loop - then 9-11 could never have happened - in fact you'll notice none of the planes were European-made Airbus - that is because Airbus planes won't fly into human targets.

This error, and the news chopper one below, and presumably many military crashes we don't hear of, cannot be controlled by ATC because ATC is not present.

ADS-B doesn't help because ADS-B is too expensive. Any solution which exceeds your budget is not a solution.

You mention radios. I can purchase a family radio with a 5 mile range terrestrial for what $49 retail. I only need the amplifier stage - the radio you mention is presumably a high-speed data packet radio with an error correcting IP stack. The radio I need is a sputnik-error pinger - much lower size, heat dissipation, and cost.

It's true they were sharing airspace - but they were not intended to fly within each others control radius. A Peer-to-peer must have a scalable resolution. Mine has - has the ADS-B?

The FAA should require first responders (including press) to carry some form of peer-to-peer system like the FLARE system we've mentioned.

I'll wager the FAA with investigate for 18 months and then declare "Pilot Error" as the cause; when It coulde be argued that the FAA's delays in adopting the best practices - largely due to entrenched GA pilots or whatnot - is the most accurate cause of these accidents.

I suggest that 911, and these kinds of event-related collisions - argue strongly in favor of a localized ATC. As a land-lubber, I resent 911 as the end result of an insistence on "potential-terrorist-in-the-loop". In my opinion, Boeing (and their investors by extension) are personally responsible for the now ~8,000 deaths, and 3 trillion in debt for delivering a product which in which the controls could be accessed by the passengers, and in which the plane could be directed to fly into buildings. Best practices _at the time_ (Airbus, and Israeli airplanes) included a separate door for pilots, and a fly-by-wire which prevents converting the airframe into a weapon.

The FAA should mandate pool coverage by helicopter - that's the only way to reduce the number of TV heli's to One. Also the Police-heli should participate in the Pool, by broadcasting their camera - The FAA can then call off all but one Heli from an incident. (But did I mention they were incompetent?)

GA are getting a sweetheart deal in that they do not share the real costs of airports on a fee-per-timeslice basis. A fee-per-passenger basis dramatically understates the opportunity costs of allocating a commercial airstrip for GA traffic. A Peer-to-peer ATC would permit more GA traffic, because it would open the skies, and dramatically lower the cost of operating a GA airfield.

You're right about Birds;
Having flown RC planes, I would appear that any UAV large enough to be a risk to others - would be large enough to warn off birds. We might do a study which shows that darker colors are more effective - since birds of prey tend to be dark.

As for the others (parachutists etc... ) There is no reason a $50 ipod-sized device could not be included.

No argument about radio range. The XTend operates at 30dBm (1W) and claims 40 mile range. A 20dBm (100mW) Zigbee chipset might be adequate, but would need to be tested. I'm just starting to look into parts costs, but expect the total to come in significantly under $50.

You have some great ideas but the FAA is legally bound by procedures and can not simply order the requirement for such systems out of hand. There is a lengthy and detailed process they must follow which can be as frustrating to some in the agency as it is to end users.
As someone deeply involved in the sUAS rulemaking process I have learned that even when the agency finally cranks to full steam ahead it will still be years before such things see the light of day.
Any such system will have to be tested and certified and liability issues addressed which will drive up costs.
Not knocking the idea, it could save lives, but if not done right it could also cause reliance on a flawed system and actually cost lives as well so it must be done by the numbers.
With regard small UAS, the real challenge remains objects to be avoided that can't or won't "ping" such as errant party balloons, sea gulls ect when flown over areas where such "encounters" could lead to things falling on people.
The type of system being discussed here has a lot of potential as a first step towards a more comprehensive sense and avoid system but for unfettered use of sUAS in some areas much more will be required.

Ira, this is bullshit.
We've had some nine medical helicopter accidents this year and you're complaining about the risks of a guidance technology. Yours is a false dichotomy; waiting for the perfect is the enemy of the good; lives are already at risk; failing to improve the system has already cost 7,000 lives in the last 6 years. (911 plus related war) That is a dismal record. New systems may have some errors - but look at the introduction of GPS navs for cars - if we had waited for the FAA to approve a TomTom, there would be one truck-mounted vendor running vmware on a mil-spec processor, it would cost $10,000 per unit, and require an additional license to use it.

The FAA, like all government, should be considered an obstacle to work around, not a necessary partner. For example Insurance carriers could (and should) offer a discount for helicopters which carry an CAS.

The real challenge is not Pingless objects: Do you have even one example of a party balloon bringing down an RC plane, or any other flying object? (Thought not) As for Birds, again - it is basically unhead of. If a plane is small enough to be attacked by a bird, it is small enough to fall without incident. Thousands of people go to baseball games and happily risk the chance of a high-intensity and completely deadly impact. Even sUAV capable of 130mph don't fall at their top speed - it takes a lot of active control to maintain speed, so uncontrolled falls, are generally lower-impact - than say a baseball.

It is not a question of perfection, but of balancing risks and benefits. Say two news choppers covering OJ simpsons white bronco collide over your house - would you prefer two 12 pound UAV to come crashing down on your house, or 3 tons of chopper coming through your roof? These are the true choices. The have or not have choice is a Logical fallacy for the kind of technical dolts a guy like Bush would appoint to run the FAA.

The only requirement for small uavs should be that they are foam with 9 inch push props (as many as you like). I challenge anyone to describe a reasonable scenario under which that uav causes more deaths than a bird might cause.


B.S. or not these issues will have to be addressed to the satisfaction of regulators and this must be done in accordance with established procedures.
I totally understand what a royal pisser this is but do we really want government officials to have the authority to issue regulations regarding matters such as this willy nilly and without public review or oversight? I think that is a frightening prospect.
As frustrating as the system is at times it looks like we will have to work with it and jump through the prescribed hoops.

Reply to Discussion


© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service