Why This Exists
Mesh networks are only as strong as their weakest gap.
Meshtastic is powerful, but ground-bound
Meshtastic devices form a mesh network by relaying LoRa radio signals from node to node, no cell towers or internet required. That makes them a genuinely resilient way to communicate when the regular grid goes down. But each node's range is limited by line of sight and low transmit power. A node sitting on a desk or a backpack can only reach so far before hills, buildings, and the curve of the earth get in the way.
Coverage gaps mean the mesh doesn't connect when it matters most
A mesh network is most valuable exactly when normal infrastructure is unavailable, during a storm, a power outage, or out in areas cell service never reached in the first place. That's also when gaps between clusters of nodes are hardest to close: nobody is deploying new ground infrastructure in the middle of an emergency, and thin, scattered adoption in any single area means node density stays low for a long time.
Altitude solves the range problem cheaply
A Meshtastic node a few thousand feet up can hear and rebroadcast between ground nodes that could never otherwise reach each other. Line of sight suddenly spans tens of miles instead of one or two. Building permanent towers to get that reach is slow and expensive. Using height that already exists is not, and there are two kinds of it. Pilots already flying can strap a registered node onboard, so a flight that was happening anyway becomes a mobile bridge. Somebody with a ridge, a tower or a tall roof can put one there and leave it, for less reach over one place and all of the time.
Your pledge buys reach, not towers
Fuel The Mesh exists to make that reach routine rather than accidental. Subscriber pledges reward operators for the coverage their nodes deliver, in the air or on a rooftop, turning a one-off relay into an ongoing, community-funded service for the city underneath. What earns a share is what the node did on the network, never the flying itself.
The aircraft are already flying.
Every day the FAA's Air Traffic Organization handles more than 44,000 flights across more than 29 million square miles of airspace. On a peak day, an average of 5,500 aircraft are enroute in US airspace every single minute, with a maximum of 5,700. None of those flights are waiting on anyone's permission or budget. They are happening regardless.
44,000+
flights handled per day by the FAA, across more than 29 million square miles of airspace
5,500
aircraft enroute every minute during peak hours on a peak day, peaking at 5,700
128 mi
the furthest confirmed Meshtastic air to ground contact on record, on stock antennas
The real argument is the radio horizon
A LoRa node on the ground is usually beaten by the curve of the earth long before it is beaten by its own transmit power. How far it can see is a geometry problem, and the geometry is extremely generous to anything that leaves the ground. The standard approximation:
horizon in miles ≈ 1.23 × √(height in feet)
Square that radius, multiply by pi, and you have the area the node has line of sight over. Because the radius grows with the square root of altitude, the area grows with altitude itself, in a straight line. Roughly 4.75 square miles of visibility for every single foot you climb. Here is that worked out, and set against a Meshtastic node on a generous 100 foot rooftop or mast:
| Height above ground | Line of sight radius | Area in view | 100 ft masts to match |
|---|---|---|---|
| 100 ft rooftop or mast | 12 miles | 475 sq mi | 1 |
| 1,000 ft | 39 miles | 4,750 sq mi | 10 |
| 2,500 ft | 62 miles | 11,900 sq mi | 25 |
| 5,000 ft | 87 miles | 23,800 sq mi | 50 |
| 10,000 ft | 123 miles | 47,500 sq mi | 100 |
| 18,000 ft | 165 miles | 85,500 sq mi | 180 |
| 25,000 ft | 194 miles | 118,800 sq mi | 250 |
| 41,000 ft | 249 miles | 194,800 sq mi | 410 |
One light aircraft at 5,000 feet has line of sight over the same area as fifty ground sites on 100 foot masts, and that is with the masts perfectly spaced on perfectly flat ground. Real terrain only makes the ground option worse. Nobody is going to fund fifty towers to cover one metro area. Aircraft cross that same area every hour of every day already.
General Aviation is not one altitude, which is why the table runs as far as it does. A training flight or a weekend cross country sits between three and eight thousand feet. A turbocharged twin works in the teens. A turboprop cruises above twenty thousand, and a business jet higher still. That is roughly seventy miles of horizon at the bottom of the band and two hundred and fifty at the top, so the same node in the same aircraft closes a gap across a county on one flight and across several states on another. Nothing about the radio changes. Only the geometry does.
This is not theoretical reach, either. The furthest confirmed Meshtastic contact between an airborne node and the ground is 128 miles, achieved on ordinary hardware with stock antennas. Run that backwards through the formula above and a 128 mile horizon needs roughly 10,800 feet, so the geometry was never the thing standing in the way. Past that height the limit stops being the curve of the earth and starts being power, antenna and noise.
The same table reads twice
Nothing in that formula asks why you are high up. The column says height above ground rather than altitude, and that is deliberate: the earth does not care whether it is holding you there or you are flying past. Every row describes a fixed site as well as an aircraft. A node on a ridge 2,500 feet above the valley it overlooks reads off the same row as an aircraft at 2,500 feet, 62 miles of horizon, and it still has it at four in the morning in February.
Which matters more than it first looks, because of when a mesh is actually needed. The storm that takes out the cell towers is the storm that grounds General Aviation. The argument above is at its weakest at exactly the moment the argument for a mesh is at its strongest, and a node bolted to a mast does not care about the weather.
Airborne Operator
Far more reach, and it moves, so one flight touches every gap along its route rather than one. Present for the length of the pass, and then not.
High Elevation Operator
Far less reach, over one place, and never absent from it. The gap either side of it is closed on the nights nobody flies, which is most of them.
Airborne operators have the bigger advantage by a wide margin. Fixed operators have the one that never goes home. Both are paid out of the same pool, on the same basis: what the node measurably did for the network.
The one percent thought experiment
This is an illustration of scale, not a roadmap. But the arithmetic is worth doing once, because it reframes what the ceiling looks like.
1% of daily flights
440
flights a day carrying a node, each opening a corridor for as long as it is airborne
1% of peak enroute traffic
55
nodes in the air simultaneously, spread across the country, moving the whole time
Take 5,000 feet, deliberately near the bottom of the band above and well under the cruise altitude of most of those flights. Fifty five aircraft at that height have line of sight over roughly 1.3 million square miles at any given instant. Matching that from the ground would take around 2,750 masts, and those masts would still be standing in the same spot an hour later, while the flights had moved on and covered somewhere else entirely.
We are not claiming one percent of US aviation is about to carry a Meshtastic node. The point of the number is the opposite: the interesting question was never how many nodes exist. It is how few flights it takes to matter.
What this does not mean
Anyone who has run a busy mesh will have objections to the above, and they are right to. Three things the numbers do not say.
- Coverage is not usable capacity.
- Line of sight means packets can physically get through. It does not mean bandwidth for everyone underneath. LoRa is a slow, shared channel, and an aircraft passing overhead does not turn into blanket service for a city. What a flight realistically buys is specific and limited: two clusters that could not hear each other can, for the length of the pass, and a gap where nobody is at all becomes briefly reachable. That is genuinely valuable, and it is not the same thing as coverage in the cellular sense.
- More altitude is not straightforwardly better.
- A node high enough to hear everything will also, if configured carelessly, rebroadcast everything, pushing all of that traffic back down onto one shared frequency. This is a known and actively discussed problem in the Meshtastic community, not a hypothetical. Meshtastic's own guidance on aggressive rebroadcast roles warns that the resulting congestion, collisions and lost packets can degrade a mesh "to the point of rendering it unusable", and tells operators to watch channel utilisation and stop once it climbs past about 25 percent. A badly placed high node is worse than no high node. An airborne one is the most extreme case of both the benefit and the failure mode, so hop limits, roles and rebroadcast settings matter more at altitude, not less. A rooftop node that hears half a county has the same duty, all day rather than for an hour. Getting that configuration right is part of what a verified operator signs up to.
- One percent is a thought experiment, not a plan.
- What we actually do is fund coverage over real cities, one at a time, starting with Orlando. The figures above exist to show why that is worth funding at all. They do not describe anything flying today, and we would rather say so plainly than let the big number do work it has not earned.
Sources and working
- Flight volume and peak enroute counts: FAA Air Traffic Organization, Air Traffic By The Numbers (June 2025 edition, FY2024 data). The 5,500 figure is the average number of IFR flights enroute per minute between 1500 and 2200 GMT on 25 July 2024, with a maximum of 5,700.
- Confirmed 128 mile air to ground contact: Meshtastic range records (206 km, LongFast preset at 915 MHz, T-Beam hardware with stock antennas).
- Congestion from aggressive rebroadcasting: Meshtastic, Demystifying ROUTER_LATE and Is LongFast Holding Your Mesh Back?, both published by the Meshtastic project.
- Horizon figures use 1.23 times the square root of height in feet, read as statute miles. That is the conservative reading of the standard radio horizon approximation: allowing for the usual atmospheric refraction, the radio horizon is nearer 1.41 times the square root of height in feet, so the true distances are further out than the table, not shorter. Areas are pi times radius squared, on flat ground, ignoring terrain, obstructions and antenna pattern. Mast equivalence is the ratio of those areas and assumes perfect placement with no overlap, which favours the ground option.