When will India Build Its Own Passenger Aircraft? | HAL CMD DK Sunil

ELI5/TLDR

India can already build fighter jets — the club of countries that can do that fits on one hand. But it still cannot design its own passenger plane. The HAL chairman says we will be assembling a 100-seater in India within five years (a Russian design, built here under license), and we might design our own from scratch in about ten. The reason it takes so long is not lack of skill — it is that aircraft are unforgiving, every part has to be tested to death before it flies, and the materials and engines need a whole industrial base that India has historically imported rather than built.

The Full Story

Why a country bothers building its own aircraft

The instinct is to think of aircraft as expensive vanity. DK Sunil pushes back. Look at the last two wars, he says. Ukraine. Iran. Even the Americans hesitated to put boots on the ground in Iran because casualties would be politically unbearable. Air power has become the way countries say something serious without committing to a full war.

It is going to become the dominant theme in all future wars and every air force, every country is now building these capabilities.

When India launched Operation Sindoor against Pakistan, the goal was a message — calibrated, not escalating. Aircraft are how you send that message.

Then there is the dependency problem. When you buy a Rafale, you do not get the keys to the car. You get the car. The mission computer — the brain that decides what the plane sees and shoots — stays under the seller’s control. Want to bolt on a new Indian-made missile? You have to ask the French. Want to upgrade the radar? Back to Paris. In a world where alliances can flip overnight — Sunil drily notes the US and NATO are currently “exchanging unpleasantries” — that is not a position a country wants to be in.

We have gone through this phase. We were manufacturing under license during the MiG days. But as we went ahead, we realized that these strings are very critical. You can’t integrate anything new without those controls.

So HAL has spent decades building the platforms it controls end-to-end: the Dhruv helicopter, the Prachand attack helicopter, the Light Combat Aircraft, the Light Utility Helicopter. When India wants to put a new weapon on these aircraft, India can.

The civilian gap

Here is the awkward bit. India is one of the fastest-growing aviation markets in the world. Every single passenger plane flying over India was made somewhere else. HAL has been so focused on military aircraft that the civilian side has barely started.

Sunil’s plan has two layers. In the short term — five years — HAL will assemble the Russian SJ-100, a 100-seater, under license in India. Manufacturing, not design. A first step. Beyond that, smaller is more interesting. India has lots of regional airports with runways too short for an A320. A 40- to 50-seater built for runways under 1.7 kilometers fits Indian geography in a way an A320 never will. Think of it as the Indian hub-and-spoke not copying the American one, but inventing its own — many small airports, many small planes.

A clean-sheet Indian-designed passenger aircraft? Maybe a decade. Possibly longer. Passenger aircraft are punishingly expensive to certify. A 100-seater costs upward of twenty-five thousand crore rupees just to design — not build, design.

What aerospace engineering actually feels like inside

The professor asks Sunil what it is like to work in a field where, unlike electrical engineering, an idea takes a decade to reach the field. Sunil’s answer is the most human moment of the conversation. He wrote a C compiler as a B.Tech project before he had even read the C book. He calls it foolhardy in retrospect. “We did it. That kind of idealism is good at that age — it makes you do wonders without knowing how complicated it is.”

The pitch to young engineers is the opposite of working at a Global Capability Center in Bangalore — those R&D-arms-of-foreign-companies that hire engineers to work on a sliver of a product they will never see whole.

At HAL, ADA, NAL — if he’s working on a control law or a rotor blade design, he sees the whole product. He knows where it fits in. That’s what gives a very good system [understanding].

He tells a story: HAL had contracted a foreign vendor to write a helicopter control law. Eight years passed, no delivery. Meanwhile a young engineer doing his PhD at IIT Kanpur on the same problem joined HAL, worked with professors, built the control law in under two years. It is already flying.

The engineering universe inside one aircraft

A modern fighter is not “an aerospace project.” It is every kind of engineering at once. The conversation rattles through the open problems, and what makes the list striking is how many of them are not aerospace-specific.

Thermal barrier coatings. Jet engine blades sit in gas hotter than the melting point of the metal they are made of. A thin ceramic coating — the thermal barrier — keeps the blade from vaporizing. India has one kind. As temperatures keep rising, new coatings are needed, and Sunil says he sees no Indian research group working on it.

Fatigue in gas turbine engines. A turbine blade flexes millions of times. Tiny cracks grow invisibly. Predicting when the crack becomes a catastrophe is the hard problem.

Thermal management of electronics. Modern avionics — the electronics in a cockpit — burn so much power that the circuit board “literally boils.” Imagine four laptop processors crammed onto one board with no fan. You need some way to suck the heat out, and Sunil admits nobody has a great answer yet.

Smart skins. A plane covered in sensors that report its own aging in real time. Imagine your car telling you, with numbers, exactly how tired its frame is.

Sensor fusion. A fighter has radars, infrared cameras, passive listening systems, all looking at the same sky and seeing different things. Merging them into one coherent picture is its own discipline.

AI in the cockpit. HAL has built an “Alexa for fighter jets” — a voice-activated control system. The problem is that Alexa works in a quiet living room. A cockpit runs at 110-115 decibels. They built it on a Hidden Markov Model — an old, well-behaved technique. Sunil is interested in neural networks but stuck on a deeper question.

How do we ensure it will behave in a certain way? This is a very big challenge. And there are no answers as of now. Nobody knows how to certify it.

That last sentence is the heart of the AI-in-safety-critical-systems problem in one line. You can make a neural network work most of the time. You cannot prove it will work next time. Regulators want proof.

Manufacturing: the part nobody talks about

Sunil tells a story about the LCA wing. It is a composite — a layered carbon-fiber structure, lighter and stronger than aluminum but more delicate to work with. The wing needs ten thousand precise holes drilled in it, and composites flex under the drill. HAL currently drills them in stages on sub-jigs, each a slow handcrafted operation. He wants to do it in one robot pass with laser precision. India does not yet have that capability.

This is the unsexy half of aerospace — methodization, the discipline of breaking a design into modular sub-assemblies so smaller vendors can build the pieces in parallel. HAL has gotten part of the way. For the LCA, one private company builds the center fuselage, another the wings, another the rear fuselage. HAL couples them. But Sunil wants the next level: vendors building entire wing assemblies with hydraulics and wiring already inside.

The deepest lesson he draws is about design itself.

The whole focus initially was getting it done. But as we went into production, we realized: it cannot be part-to-part specific. If a wing is not interchangeable, it has to be matched to that specific aircraft. It becomes a problem in the field.

So newer HAL projects are built straight from 3D models — design, manufacture, and inspect from the same digital file. Engineers call this Design For Manufacturing. Sunil says they now “drill it into the juniors” from day one: talk to the manufacturing people before you finalize the design, not after.

The thin margin problem

Civil engineering builds bridges with a safety factor of two or three — if a beam needs to hold ten tons, you design it for twenty. Aircraft cannot afford that. Extra metal means extra weight means less fuel means less range. The factor of safety in aerospace is closer to 1.1.

Sunil admits even his engineers want to play it safe — they secretly use 1.5 when they should use 1.1. The result is an overweight plane. The discipline is to trust the math, the testing, the simulation. Hundreds of physical specimens. Reviews at every gate — Preliminary Design Review, Critical Design Review, sub-system tests, full structure tests. Years pass. By the time the aircraft flies, nothing inside it has not been broken on the ground first.

He thinks the future is less testing, more simulation — the digital twin. Build a perfect virtual copy of the aircraft, run a thousand simulated flights overnight, and only build the physical thing when the math is clean. India is not there yet. The current loop, he confesses, is “fly-fix-fly.” Pilot flies, pilot comes back with a list of bugs, engineers fix, repeat.

The thing nobody saw coming: the supply chain

The professor frames it as a luxury — every aircraft maker globally has a four-year order book. Sunil’s response is a quiet warning.

Some of the mills which produce these special alloys — they’re fully booked for the next five years. By the big engine manufacturers.

The Ukraine war made Europe ramp defense spending to three to five percent of GDP. Revenge tourism after COVID made airlines order planes by the hundred. Result: the world’s titanium mills, nickel alloy foundries, and component plants are sold out for years. India cannot just buy its way in. The professor asks, half-jokingly, whether HAL could build aircraft faster if the parts simply arrived. Sunil is blunt: yes. Much faster.

So India is now spending money with companies like MIDHANI to make titanium alloys, nickel-based superalloys, and high-temperature castings inside the country. This is where the material scientist’s career opens up — a generation of underinvested chemistry, suddenly strategic.

Drones and the new shape of war

The recent Iran-Israel exchange showed what a few thousand drones can do. India has a thriving startup ecosystem in the 10-20 kilogram class — the small reconnaissance drones, the loitering munitions. Sunil is happy to let the startups have that market. HAL is going the other way: bigger.

HAL’s flagship drone is the Warrior — a 2.5-ton unmanned combat aircraft with its own gas turbine, designed to fly alongside manned fighters as a “wingman.” Later versions will go to three tons, five tons, carrying bombs and missiles.

A clever experiment: HAL took a Kiran trainer — a two-seater where the trainee sits beside the instructor — pulled out the trainee seat, and put flight computers in its place. They call it Optionally Manned Combat Aircraft. The plan is for a pilot to take off, hand control to the computer mid-flight, then later pull the pilot out entirely. A staged way to teach a fighter to fly itself.

Sunil flags two specific opportunities for Indian startups. First, GPS-denied navigation. Operation Sindoor revealed that many imported drones failed when GPS was jammed — and Iran has gotten very good at jamming GPS. A drone that navigates without satellites is a strategic capability. Second, small gas turbine engines. Batteries do not have the power density for long-endurance flight; you need a turbine. India needs a whole family of small engines — turboprop, turbojet — in the 1 to 10 kilonewton range.

The career argument

The professor asks what mindset a young engineer needs. Sunil’s answer is patience, packaged as a metaphor.

I always ask people to put a large capacitor — wait for a little time. And then you will find that it becomes more and more interesting.

A capacitor stores energy and releases it later. Aerospace is the same. Other disciplines give you a payoff in three years. Aerospace gives you nothing for five and a career for forty. The thing you build will outlive you. The technology gets reused across decades, across platforms. Your grandchildren can point at it.

The risk is the opposite — boredom in a narrow MNC job where you work on a slice of a system you will never see whole.

Key Takeaways

India will assemble a 100-seater civilian aircraft (Russian SJ-100) under license within five years; a clean-sheet Indian passenger aircraft is roughly a decade away.
The strategic case for building your own fighter is not patriotism — it is that imported aircraft lock the mission computer and flight control software with the seller. Integrating new weapons requires their permission.
The club of countries that build their own fighter jet fits on fewer than two hands; India is in it thanks to HAL.
Designing a 100-seater passenger aircraft costs upward of 20,000-25,000 crore rupees — before manufacturing.
India’s airport geography (short runways, ~1.7 km) makes a 40-50 seater more useful than a copy of the A320; this is the regional sweet spot.
The bottleneck in aerospace is not engineering talent but materials — titanium alloys, nickel-based superalloys, thermal barrier coatings, composites. India under-invested for decades because importing was easy.
Globally, special-alloy mills are sold out five years deep. India can no longer buy its way in.
Modern aircraft electronics burn so much power that boards “literally boil”; thermal management is a wide-open problem.
Hidden Markov Models work in safety-critical voice control; neural networks do not — because nobody knows how to certify them. The certification problem, not the capability problem, gates AI in aviation.
HAL writes Design For Manufacturing (DFM) into projects from day one — engineers must talk to manufacturing before finalizing designs, because retrofitting modularity is ruinously expensive.
Aerospace safety factors are around 1.1 (vs ~3 in civil engineering); the cost of conservatism is weight, the cost of weight is range, the cost of range is mission.
The future of aerospace iteration is digital twin simulation, not physical testing — the current “fly-fix-fly” loop is too slow.
HAL is building optionally-manned aircraft by converting trainer aircraft — a staged path to fully autonomous combat aircraft.
A young HAL engineer (PhD from IIT Kanpur) built a helicopter control law in under 2 years that a foreign OEM had failed to deliver in 8 years.
Small drones (10-20 kg) are a startup-friendly market; large UCAVs (2.5+ tons, gas-turbine-powered wingmen) are HAL territory.
GPS-denied navigation is a major capability gap exposed by Operation Sindoor — many Indian drones failed when GPS was jammed.
A “UAV alliance” of non-Chinese drone controllers is a strategic ask — current controllers come from China with unknown backdoor risk.
Air power has become the politically preferred way to send a military message because troop casualties are politically unbearable — applies equally to Ukraine, Iran, and India’s Sindoor operation.

Claude’s Take

Sunil is exactly the kind of CMD interview that should be done more often and almost never is. He talks like an engineer, not a press release. The whole conversation has the texture of someone who actually solves problems for a living — he can name the specific bearing mount that vibrates wrong on a gas turbine, he knows the decibel level inside an LCA cockpit, he knows that ten thousand holes in a composite wing flex when you drill them.

The honest moments are the best ones. “Today, even today, I’m telling you, we still don’t have good models” of thermal behavior in electronics boxes. “Nobody knows how to certify” neural networks for aircraft. These are not the lines a defense PSU chairman is supposed to say on camera. They are the lines that make you trust everything else he says.

The strategic picture he paints is more sober than the average aerospace-in-India conversation. The civilian aircraft program is realistically a decade off. The supply chain crisis is severe — you cannot just announce you will build aircraft when the world’s titanium mills are booked through 2031. The drone gap with China is real, and not closing fast. None of this is doom-mongering; it is just an inventory.

What is missing is any discussion of the elephant in the room — HAL’s production delays on the Tejas LCA, the IAF’s repeatedly delayed orders, the actual gap between promised and delivered. The professor does not ask, and Sunil does not volunteer. That is the only place the conversation pulls its punches. Still, the rest is rich enough that an 8 feels right. This is a deeply substantive interview that rewards careful listening — the policy-bait headline buries an hour of actual engineering.

When will India Build Its Own Passenger Aircraft? | HAL CMD DK Sunil | Episode 54