Electronic Conspicuity Monitor

A live measurement of how light aircraft make themselves visible electronically: ADS-L by radio and, as its mobile variant takes shape, by phone, alongside FLARM, FANET and the phone apps that already share positions. Each answer is given for light aviation as a whole and then for each kind of aircraft, because rules that work for one kind can fail another.

Also here: why this monitor exists, the method with every rule and its date, the field notes, on what the feed carries besides aircraft: balloons on a test bench, weather stations, drones, airliners typed as paragliders; and how light aviation flies.

How this page works. The page is computed live from the Open Glider Network feed. The map shows positions as they arrive and refreshes every five seconds. Every packet is also measured as it arrives; the totals are written every 15 minutes and the figures on this page are recomputed from them every ten minutes, so they move a little from one visit to the next. The answers add up everything since 4 October 2026, when version 1 of the method took effect, except the time without signal, counted since 06:07 UTC on 6 October 2026 under the rules of that day; the coverage share, the map of squares and the table of app intervals add up the current month and the previous one, and the two charts that count aircraft show the latest complete month. Each card says how many hours or predictions it rests on. The rules behind every figure were published before the data was collected, on the method page, with every later change dated; the source code is public, and every figure can be downloaded as CSV or JSON to be checked. Why the monitor exists and how it measures, in plain words: why this monitor, and how it works.

Questions 2, 3 and 6 compare the channels by the time an aircraft is without signal: the part of each silence beyond the interval its source keeps by design, plus 10 seconds. How often a source should send is question 4.

Feedback, corrections and questions are welcome: the address is in the box at the end of the page.

Live

1 · Should ADS-L carry the turn rate?

Gliders, paragliders and hang gliders spend much of their flight circling in thermals, often several in the same one. Anyone predicting where such an aircraft will be a few seconds later, to avoid it, needs to know that it is turning. Either the transmitter says so, or the receiver works it out from the positions it has already received.

FLARM and the OGN tracker send their turn rate; ADS-L has no field for it, so an ADS-L receiver has to work it out from the last two packets it heard. When every packet arrives the two are a second apart and the derived turn is good. When packets are lost, as they are when the pilot’s body shields the radio (measured under question 2), the last two can be many seconds apart: in 8 seconds a paraglider in a thermal turns through a third of a circle, in 16 more than half, and the turn worked out from them goes wrong. The packet that does get through still carries the turn rate.

The chart measures this on real tracks that send every second, by pretending the packets in between were lost: the receiver keeps the start packet and the one 2, 4, 8, 16 or 32 seconds before it, the last being about the interval at which a phone app sends. It then compares, at the same instants, an arc with the turn the receiver derives from that pair, an arc with the turn rate the start packet carried, and a straight line. The error is the distance between each prediction and the position that then arrived.

Free flightA paraglider pilot’s body shields the radio for part of every circle (measured under question 2), so paragliders lose more packets than aircraft whose antenna has a clear view; the card above gives, for each kind of aircraft, how often a gap of more than 16 seconds occurs. The longer the gaps, the more a turn rate in the packet is worth.

2 · Radio or phone: which keeps aircraft visible to the network?

U-space services, drones, rescue teams and tracking maps see aircraft through a network: ground receivers for radio, the internet for phones. The question is which of the two keeps hearing an aircraft for more of the flight, and whether the answer is the same for every kind of aircraft.

What is measured: time without signal. Some apps send a position once a minute by design, and in that minute a glider covers two kilometres whatever the network does. How far behind an aircraft falls therefore depends on how often each app chooses to send, which is a property of the app. To compare the channels themselves, the page counts only the time without signal beyond the interval each source keeps by design, plus 10 seconds: one second for FLARM and ADS-L, two for OGN trackers and PilotAware, fifteen for FANET (whose specification slows it down where many fly), and for each app the interval in the table under question 5. Sources whose interval is not known are left out. How often apps should send is question 4. The first chart gives that share of flying time for each kind of aircraft, by phone app and by radio. The second shows what a map displays, source by source: the share of time with the position more than 300 m, 1 km or 3 km behind the aircraft, which includes each source’s own cadence.

This is visibility to the network. Between two aircraft close to each other, two radios hear each other directly, even where no ground receiver does, and a phone does not; that case is outside what this feed can show.

Free flightA paraglider pilot hangs in a harness below the wing, with the instrument on the harness or in a pod in front of the chest, so the pilot’s own body stands between the radio and part of the sky. A phone in a pocket has the same body around it, but it talks to whichever mobile cell is in view.

Why radio loses paragliders. The pilot’s body can be measured directly. A paraglider circling in a thermal turns through a full circle every 20 to 30 seconds, so its transmitter points at a ground receiver from every side in turn, equally often. An antenna that radiated equally well all round would be heard equally often from every side; where the body shields it, packets go missing. The chart counts the FLARM packets received from circling aircraft by the direction of the receiving station relative to the aircraft’s heading, for free flight, gliders and powered aircraft. Gliders and powered aircraft are the comparison, and an imperfect one: FLARM itself says the human body attenuates its signal, and in most gliders the antenna sits on the instrument panel with the pilot right behind it. The chart therefore compares three ways of carrying a transmitter: on a paraglider pilot’s harness, in front of a glider pilot, and in a powered aircraft, where the antenna may be on the panel, on the canopy or under a metal fuselage, and where circling is mostly the turns of a circuit.

What the chart sees, and what it cannot. The receiving stations are on the ground, below the aircraft, so the chart shows the signal that leaves the transmitter downwards. A station 5 km away and 1,000 to 2,000 m lower is seen 10 to 20 degrees below the horizon; at 20 to 40 km the angle is 2 to 6 degrees. A large share of the packets is received by stations within 5 km, where the angle is steepest. While circling the aircraft is also banked, by 20 to 45 degrees, so a station on the inside of the turn lies well below the pilot’s own horizontal plane and one on the outside lies above it. Another aircraft at the same height sits in that horizontal plane, where the pilot’s body and the harness stand in the way differently. The chart therefore measures the shielding towards the ground network and towards aircraft flying lower. Towards aircraft at the same height it may be stronger or weaker, and no ground receiver can measure it.

3 · Does mobile coverage fade with height above the ground?

A phone app needs mobile coverage, and mobile networks are built for people on the ground. Where coverage is missing, the app’s positions stop arriving: anyone watching still sees the aircraft at the last place it was heard, while it flies on. The charts show how often that happens at each height above the ground, for apps and for radio.

The first chart shows the share of flying time without signal, the measure of question 2, which leaves out the interval each source keeps by design. A distance would mix in the speed as well: a kilometre is a minute and a half of silence for a paraglider and 15 seconds for an aircraft at 240 km/h, and gliders and powered aircraft fly faster high above the ground than in the circuit. It is given by height above the ground at the moment the aircraft was last heard. That share counts only the gaps that end, when the aircraft is heard again; an aircraft lost for good adds nothing to it. The second chart counts those too: how many times per hour flown an aircraft went silent for more than two minutes at each height. Near the ground that includes landings; high above it nobody lands, so a disappearance there is lost coverage. Heights come from a terrain model of Europe at 15 arc-seconds. The bands follow the limits stated by mobile operators and quoted in a 2021 desk study for EASA: coverage reliable up to 300 m, patchy up to about 1,000 m, and a link lost between 600 and 1,200 m in trials the study cites.

How the height above the ground is worked out, and how far it can be off. It is the aircraft’s altitude in the packet less the elevation of the ground below it, taken from the NOAA ETOPO 2022 terrain model. The model’s cells are about 460 by 310 m in the Alps; the squares of the map in question 6 are a different, much coarser grid. The ground under the aircraft is interpolated between the four nearest cells, which follows a slope across a cell; on the steepest ground of the Alps that halves the error or better, to a median of about 30 m. Relief smaller than a cell stays, so close to a ridge, where a pilot may soar 50 m from the slope, the model can still be off by some tens of metres to over a hundred, and a few segments near a slope land in the wrong band. An error of that size moves segments that lie near a band boundary; it cannot turn a band where apps rarely lose aircraft into one where they often do. The altitudes from radio sources are above sea level, like the terrain model: FLARM and ADS-L transmit a height above the GPS ellipsoid, and the Open Glider Network receiver converts it, while FANET and OGN trackers send sea-level altitude directly. A few receivers skip the conversion and report about 45 m too high. For phone apps nobody documents which altitude they forward; one that passes on an Android phone’s raw height above the ellipsoid would read about 50 m high in Europe. That is still being checked, by comparing pilots heard through an app and through a radio at the same time.

These figures describe the mobile networks of today, built for people on the ground. Aerial use of those networks has been harmonised in Europe since 2022 (ECC Decision (22)07); the Commission has asked CEPT to prepare the use of 5G base stations to sense flying objects in the 3.6 GHz band (ISAC), which may also bring coverage higher; and direct-to-device satellite services are on their way. Coverage aloft is expected to improve, and this measure will show by how much.

4 · Every so many seconds, or every so many metres?

Requirements for U-space are written in seconds: an update every 3 to 6 seconds. A rule in seconds asks every aircraft to transmit equally often, whatever its speed, so it is strict on slow aircraft and lenient on fast ones. A rule in metres asks an aircraft to transmit once it has moved a set distance, which keeps every aircraft’s position equally fresh on a map.

For each source and kind of aircraft, the chart sets two readings of the same flights side by side: the share of airborne time that a 6-second rule would count as stale, and the share during which the position on the map had in fact fallen more than 300 m behind the aircraft. Where the first bar is long and the second short, the rule in seconds condemns an aircraft that anyone watching could still see; where it is the other way round, the rule passes an aircraft that has already gone.

Free flightA paraglider flies at 30 to 40 km/h and its instrument, often a phone, runs on a small battery for a whole day. A 6-second rule makes it transmit ten times more often than its movement needs, and still calls it stale when it sends every 150 m and is never more than 150 m from its last position.

5 · Free flight cannot be made to carry a device. How much of it happens where phone apps work?

A powered aircraft or a glider flies from an airfield, where equipment can be required and checked. Paraglider and hang glider pilots, some 200,000 active in Europe, about 110,000 of them in the federations that make up EHPU, carry their aircraft up a mountain in a rucksack and take off from a slope; nobody can check what is in the rucksack, and a mandate that cannot be enforced will not be followed. What they all carry is a phone. What matters is how much of their flying takes place where a phone keeps them visible.

A square of the map counts as covered when phone apps logged at least two hours of airborne time there in the current and previous month and had signal for at least 95% of it. Each app sends its position at its own pace, some every second, some once a minute, some more often the faster the aircraft moves, so only the time beyond the interval an app keeps by design counts as time without signal. The flying time of every paraglider and hang glider, whatever it carries, radio included, is then split by how well apps did in its square. The answer is the share of the time in squares with enough app traffic to judge. The time in the other squares is shown apart, and counting it as not covered gives a lower bound for all free flight.

6 · Where do aircraft disappear?

Each square is about 25 by 20 km. Its colour shows for how much of the flying time spent there the aircraft was without signal on the channel chosen above the map, with the same measure as question 2, which leaves out the interval each source keeps by design. Where phone apps are green and radio is red, an app is the better way to be seen there, and the other way round.

How much each kind of aircraft flies

Collecting.

Adoption of ADS-L and of the other systems

Which systems each kind of aircraft carries, as the network hears them. Powered aircraft and gliders fly from airfields and can be asked to fit equipment; free flight cannot, and what it carries is what pilots choose.

The phone share is a floor, and lowest for free flight. Only apps that forward positions to OGN are counted, and few of the apps used in free flight do so today; a pilot tracked live by any other app appears here only if a radio device is also heard. The free-flight share by phone is therefore the most understated figure on this page.

Each transmitter identifies itself with an address. Most use a fixed one, either the aircraft’s ICAO address or one assigned by FLARM or OGN, and each of those is counted once per month. Some transmit a random address that changes at every power-up or more often, so the same device can show up several times in a month: those are drawn as a lighter band on top, and the true number of devices lies somewhere inside it. A device flying out of range of every receiver goes uncounted.

The current month is still filling up, so it is drawn dashed and its low value is no sign of a drop. Light aviation is also seasonal, and winter months will dip whatever happens to adoption. The fair comparison is the same month one year apart, which appears as a grey line once there is a year to compare with.

Radio and phones

Besides ADS-L, the Open Glider Network relays FLARM, FANET and the positions that several phone apps send over the internet, among them SafeSky, SeeYou Navigator and VarioVoice. Those apps use protocols of their own today; the pilots using them are the most likely users of ADS-L Mobile, the variant of ADS-L carried over the mobile network and one of the three means accepted inside U-space, once its specification is finished. The map above can show them, and the chart below counts every source in the feed.

Radio and internet figures measure different things. A radio count stops where the receivers stop. An app count has no coverage limit, but it includes only the users who chose to share their position, and only for the apps that forward to OGN at all. Neither is the number of pilots. ADS-B, mostly airliners, is left out of the chart.

The systems

Inside U-space, the airspace where drone traffic is managed by digital services, three means of being electronically conspicuous are accepted: ADS-B, ADS-L over radio (SRD860) and ADS-L Mobile. The other systems below are the ones pilots already carry. An open standard is one that anyone may implement without a licence.

Method and data

The full method, with every rule, the decisions taken and the date each was fixed, is on its own method page. In short: every figure comes from the public APRS feed of the Open Glider Network, which carries what its volunteer ground receivers hear by radio and what apps and platforms forward over the internet. A packet counts as radio when it carries the reception figures a receiver adds, signal-to-noise and frequency offset, and as internet otherwise. Each source is identified by the destination field of the packet, as listed in the OGN protocol repository. What the feed turned up along the way, from a weather balloon flying as a paraglider to an aircraft over Germany drawn in the Gulf of Guinea, is collected in the field notes.

Every count is a floor. A device out of range of every receiver, or an app that does not forward to OGN, is not seen. Random addresses can count one device several times, and they are shown apart for that reason. The database keeps one row per device and month for the counts, and daily totals for the visibility measure. The raw feed is kept on the server for seven days, so that a change to the method can be tested on real data before it goes live, and is then deleted; nothing else stores a track or a position. The live map refreshes every five seconds and the statistics every ten minutes.