5G unblocked BVLOS in a weekend. The paperwork took eighteen months.

In 2022 we ran a 5G-enabled UTM trial with Vi. Our stack — built on top of Aerobridge, the open-source DGCA-compliant management server for drone operations — talked to drones in BVLOS over the 5G link. It worked. The technical case closed in a weekend. The regulatory case took the better part of two years.

That's the headline.

What 5G actually buys you

Most of the public discourse on 5G drones fixates on latency. That's the wrong axis. For a multi-rotor flying a survey or a delivery, the closed-loop control runs onboard — you don't need sub-10ms RTT to a cell tower to keep a drone in the air.

The interesting wins live elsewhere:

• Uplink throughput. A 4K live feed off the drone is the use case that breaks LTE in any meaningful number of cells. 5G fixes it.
• Cell capacity at events. When ten drones are operating over the same square kilometer, the LTE cell gives up before any of the drones do. 5G holds.
• Slicing, in theory. Guaranteed-bitrate slices for telemetry are the right design. In a 2022 trial environment they were also more story than substance — we engineered around best-effort, not toward a QoS we couldn't depend on.

None of that gets you BVLOS.

The actual bottleneck

BVLOS — flying a drone past the operator's line of sight — is a regulatory problem dressed up as a technical one. In India that means DGCA approvals, NPNT compliance, conditional flight permissions, segregated airspace corridors, real-time deconfliction infrastructure the regulator trusts, and, for any serious operation, a relationship with the local airspace authority that takes longer to build than any piece of software in the stack.

The network unblocked the demo. It did not unblock the operations.

You can stand up a 5G UTM stack in an afternoon now. You will not get permission to do anything interesting with it that fast.

What I'd tell anyone starting today

If you're building drone operations on 5G in 2026, the order is: regulator relationship first, conformance second, comms layer last. The cellular industry has spent four years building drone-grade radio. The state has not spent four years building drone-grade approvals. You don't get to skip that step.

The network is the easy part.