BVLOS Drone Operations in India (2026): What the Rules Allow, What an Exemption Demands, and How to Close the Link

A procurement note asks for a drone with "50 km range". The airframe is rarely the hard part: plenty of fixed-wing and hybrid VTOL aircraft built in India can fly 50 km and back. What decides whether that range is usable is two things the specification sheet usually leaves out. The first is whether a civil operator is legally allowed to fly beyond visual line of sight (BVLOS) at all. The second is whether the command-and-control (C2) link actually closes at that distance once terrain, antenna height and the aircraft's own altitude limit are accounted for.

This guide covers both, as of October 2026, for integrators, research institutions, infrastructure owners and defence users. For the baseline civil rules (type certificate, UIN, remote pilot certificate, airspace zones) read our DGCA drone rules compliance guide first; this post assumes that baseline.

Where BVLOS stands legally in India (October 2026)

The Drone Rules, 2021 neither define BVLOS nor create a standing permission for it. Visual line of sight appears only in Rule 3's definition of a model remotely piloted aircraft system, and the green, yellow and red zone permissions in Rules 21 and 22 govern where you fly, not how far you can see the aircraft.

In practice, every civil BVLOS operation so far has run on a conditional exemption issued by the Ministry of Civil Aviation. Under the current rules that power sits in Rule 48, the "general power to exempt", which lets the Central Government exempt a person or class of persons from the rules, wholly or partly, subject to conditions in the order. Before August 2021 the equivalent power was Rule 70 of the short-lived UAS Rules, 2021.

• May 2019: MoCA invites expressions of interest for BVLOS experiments and later sets up the BVLOS Experiment Assessment and Monitoring (BEAM) Committee.
• May 2021: PIB reports that 20 of 34 applicant consortia received conditional exemptions for experimental BVLOS flights, valid for one year or until further orders. The stated purpose was to gather data for a future BVLOS framework.
• 2021: further conditional exemptions for specific BVLOS trials, including the Government of Telangana's Medicine from the Sky trials (May 2021), a Meghalaya health-supply route for the Department of Health and Family Welfare (July 2021) and CSIR-NAL (September 2021).
• September 2025: the draft Civil Drone (Promotion and Regulation) Bill is published for consultation. It would replace the Drone Rules and, among other changes, introduces imprisonment of 3 to 36 months for some offences. Check its status before relying on it; at the time of writing it had not replaced the 2021 rules.
• August 2026: Civil Aviation Minister K. Rammohan Naidu tells the Andhra Pradesh Drone Industry CEO Round Table at Orvakal that "by the end of this year, we will have a refined certification framework with enablement of BVLOS certification".

So today no type certificate endorsement, licence class or DigitalSky permission by itself authorises a civil BVLOS flight. Until the promised framework is published, you are asking MoCA for a mission-specific exemption, and the timeline should allow for that.

What a BVLOS exemption actually asks for

The published exemption orders are the best guide to what the eventual framework will demand, because they show what MoCA and DGCA have already insisted on. The July 2021 order for the Meghalaya health-supply route is a good example, and its conditions read like a checklist an engineering team can design to:

• A fixed route between named sites (Mawkyrwat CHC to Ranikor CHC), not open-area flying.
• Maximum 400 ft AGL; anything higher needs Airports Authority of India clearance.
• A 15% flight-time energy reserve on every sortie.
• Daylight only, with visual meteorological conditions at take-off and landing sites.
• Security clearance from the Ministry of Home Affairs for the airspace, plus Ministry of Defence and local administration clearance before trials.
• A DGCA-approved BVLOS operating SOP and a separate ATC and IAF coordination SOP, including contingency procedures for C2 link loss.
• A hazard identification and risk management workshop before the first flight.
• Insurance against injury, death or property damage to uninvolved parties, and an indemnity in favour of the Government of India.
• Remote pilots with valid training and a safe VLOS record.
• A proof-of-concept report to MoCA and DGCA afterwards, and situational awareness data to ATC and IAF units on request.

Notice what is not on that list: no minimum range, endurance or sensor. The regulator cares about where the aircraft goes, what happens when something fails, and who else in the airspace knows. The industry asks made at Orvakal in August 2026 point the same way: risk-based certification, dedicated flight-test environments and clear operating rules. Teams that already write their safety case in the style of the JARUS SORA method used by EASA and several other regulators will find it easier to adapt to whatever India publishes.

Engineering the C2 link: the number that limits most BVLOS plans

Most "long range" claims quote line-of-sight range from a datasheet. A real BVLOS route has to close three separate budgets: free-space path loss, Fresnel-zone clearance over terrain and trees, and the radio horizon set by antenna heights. At the 400 ft (about 120 m) ceiling that civil exemptions have used, the third one bites earlier than most people expect.

Worked example: a 20 km corridor

Free-space path loss in dB is 20·log10(d in km) + 20·log10(f in MHz) + 32.44. At 20 km that gives about 117 dB at 866 MHz, 126 dB at 2.44 GHz and 134 dB at 5.8 GHz. Every doubling of distance adds 6 dB, so 40 km costs another 6 dB at every frequency.

Take a typical digital video link at 5.8 GHz: 27 dBm transmitter on the aircraft, 2 dBi airborne antenna, and a receiver that needs roughly -88 dBm for a few megabits per second. With an 8 dBi ground omni, received power at 20 km is about -97 dBm, so the link fails before any fading margin. Holding a 10 dB margin, the same setup reaches only about 2.3 km. Swap the omni for a 23 dBi tracking antenna and received power rises to about -82 dBm: 6 dB of margin at 20 km, and a 10 dB margin out to roughly 13 km. That is why serious long-range video uses a tracking antenna, and why the range on a product page should always state the ground antenna used to get it.

Now the telemetry and command link. A narrowband sub-GHz modem at 866 MHz with 20 dBm output, a 3 dBi ground antenna, a 0 dBi airborne whip and around -110 dBm sensitivity arrives at about -94 dBm at 20 km, leaving roughly 16 dB of margin. Low data rate, but it is the link that has to survive, because it carries the commands that bring the aircraft home. Two cautions for India. Telemetry radios sold as "900 MHz" are usually built for the US 902 to 928 MHz band, which is not licence-exempt here; the Indian licence-exempt sub-GHz allocation is 865 to 868 MHz under G.S.R. 853(E), with power limits covered in our LoRa frequency bands guide. And on the uplink, a high-gain ground antenna multiplies transmitter power into EIRP, which can exceed the limit for a de-licensed band even when the transmitter itself is compliant. Many designs keep the dish receive-only for video and carry command on a modest sub-GHz uplink.

Fresnel clearance and the radio horizon

The first Fresnel zone radius at the midpoint of a 20 km path is about 16 m at 5.8 GHz and about 42 m at 866 MHz. You want at least 60% of it clear, which is 25 m at 866 MHz, plus the height of whatever stands under the path: trees, buildings, a ridge. Earth bulge adds another 6 m or so at the midpoint of a 20 km path. With a ground antenna on a 10 m mast and the aircraft at 120 m AGL over flat ground, the straight line between them sits about 59 m above the surface at the midpoint. That clears 25 m of Fresnel zone plus 20 m of trees. Let the aircraft descend to 60 m for a closer look and the midpoint clearance drops to about 29 m, which does not, and the link fades exactly when the operator is concentrating on the payload.

The radio horizon for a 10 m mast and an aircraft at 120 m is about 58 km in standard atmosphere; at 60 m it is about 45 km. In flat terrain that is the hard ceiling for a single ground station at legal altitudes. In hills it is irrelevant, because a ridge cuts the path long before the horizon does. Routes in the Northeast, Uttarakhand or the Western Ghats almost always need a relay: a second ground station on high ground, an airborne relay, or a tethered UAV holding a relay radio aloft. A tethered aircraft is powered from the ground and stays within its tether length, so it raises the antenna without raising a BVLOS question of its own.

Satellite and cellular links

Geostationary SATCOM removes the horizon problem but adds weight, power and roughly 500 to 700 ms of round-trip latency: fine for supervising an autopilot, not for hand-flying. Cellular C2 is patchy at 120 m AGL because base-station antennas are down-tilted, and airborne use should be confirmed with the network operator and in your exemption application rather than assumed.

Lost-link logic is the core of your safety case

Regulators ask about C2 link loss first because that is the failure every BVLOS aircraft will meet. Write down, and test, what the aircraft does in each case:

• Short dropout (a few seconds): continue the planned leg or hold, without operator input. Configure the timeout to the longest dropout you measured in route surveys, not the autopilot default.
• Sustained loss: climb to a pre-surveyed safe altitude within the approved ceiling, then fly to the nearest rally point or back to launch along a pre-cleared track, not a straight line that may cross a red zone or a village.
• Loss of GNSS as well as C2: a defined flight-termination behaviour inside a defined area. For a fixed-wing or hybrid VTOL aircraft that may mean a parachute or a controlled descent into a surveyed field.
• Energy: compute the reserve for the worst case, which is a lost link at the far end of the route into a headwind, then add the regulator's margin (15% of flight time in the 2021 order) on top of that, not inside it.
• Geofence: a hard boundary in the flight controller that matches the exempted corridor, so a navigation error cannot carry the aircraft into restricted airspace.

Rule 12 of the Drone Rules lets the government notify mandatory safety features, naming "No Permission No Takeoff" hardware and firmware, a real-time tracking beacon and geofencing as examples. Building those into a BVLOS aircraft now costs little and answers questions an approving officer will ask anyway.

Detect and avoid at low level

Below 400 ft the traffic that matters is helicopters, paragliders near hill sites, spraying drones and birds. Much of it carries no ADS-B, so an ADS-B receiver alone is incomplete. Workable layers:

• Procedural separation: a route away from known helicopter corridors and helipads, a NOTAM or AAI coordination where the exemption requires it, and a direct line to the nearest ATC and IAF unit.
• Ground observers at points along the route who can call traffic, which also gives you independent link-quality reports.
• ADS-B In on the ground station and, if weight allows, on the aircraft, accepting that it only sees equipped traffic.

Choosing an aircraft for BVLOS work

Range is the least useful number. What separates aircraft for BVLOS work:

• Hybrid VTOL or fixed-wing: best for linear corridors (pipelines, power lines, borders) and supply drops, because cruise efficiency gives the energy reserve room to breathe. Needs a clear vertical take-off and landing area at both ends but no runway.
• Multirotor: simple and good for point inspection, but with endurance usually under an hour the reserve rule eats much of the useful range.
• Tethered multirotor: unlimited endurance on ground power, fixed location, no BVLOS question. Often the right answer for perimeter surveillance and for lifting a relay radio.
• Multi-band datalink: separate command and video links, with frequency hopping on command, is the baseline. Ask the supplier which bands, at what power, and with what ground antenna the quoted range was measured.
• Logging: flight and link-quality logs you can export for the proof-of-concept report.

For corridor work, our VTOL X1 hybrid VTOL sits in the first class. For shorter tactical ISR within visual or near-visual range, our long-range surveillance drone carries EO and thermal sensors. Its radio range is not what a civil operator can use without an exemption: a small multirotor is hard to see beyond a few hundred metres, so under VLOS rules eyesight, not radio, sets the limit.

A practical sequence for a first BVLOS programme

• Define the mission as a route, not a capability: start and end points, corridor width, altitude band, time of day, payload.
• Survey the route for link margin before applying: fly it VLOS in segments, or drive it with a test transmitter on a mast, and log signal strength.
• Write the SOP and the contingency procedures, including lost link, lost GNSS, fly-away and emergency landing sites, and run a hazard workshop with the people who will fly.
• Line up the stakeholders the 2021 orders named: MoHA for security clearance, AAI if any segment goes above 400 ft, the nearest ATC and IAF units, district administration, and an insurer that will write third-party cover for the operation.
• Apply to MoCA for a conditional exemption under Rule 48, or under the new framework if it has been notified by the time you read this, attaching the route survey, SOPs, aircraft type certificate and UIN, pilot certificates and insurance.
• Fly, log everything, and report. That record is your best evidence when the certification framework arrives.

Research institutions can do much of the groundwork earlier. Rule 42 lets R&D organisations, educational institutions, recognised start-ups and manufacturers fly without a type certificate, UIN or prior permission, but only within their own premises or an open green-zone area under their control. That is enough to test lost-link behaviour, geofences and relay architectures before any public route is involved, and it is why a campus drone lab with a controlled outdoor test area is worth more for BVLOS research than an indoor flight cage alone.