Chapter 249 - Chapter 238: The Weight of Range

Chapter 238: The Weight of Range

The desert at Pokhran in the third week of September had the quality of having decided, finally, to stop pretending it was anything other than what it was. The monsoon's last gesture had retreated north in the final days of August, leaving behind a specific kind of dryness that was not the dryness of summer — which carried heat as its primary characteristic — but the dryness of post-monsoon Rajasthan, which carried clarity. At six in the morning, with the sun still low enough that its angle came in flat and horizontal across the scrub and the salt flats, you could see in any direction for distances that felt, to a pilot standing on the apron of the Pokhran Air Force Range's forward operating base, like more horizon than any landscape had earned.

Wing Commander Arjun Veer Bhatia had been standing on that apron since five-forty, because Wing Commander Bhatia did not believe in arriving at a briefing at the time the briefing was scheduled. He believed in arriving early enough that the briefing, when it started, would not be able to contain anything he hadn't already had time to think about. He was thirty-eight years old, a fighter pilot for seventeen years, currently assigned to the Tactics and Air Combat Development Establishment at Air Force Station Jamnagar, and he was holding a steel cup of tea that a corporal had produced from somewhere in the operations building and that had gone from scalding to merely hot in the twenty minutes he'd been standing in the morning air looking at what the pre-dawn sky was doing with the first light.

It was doing, in his private assessment, something very good. The cloud base was at approximately 12,000 feet according to last night's met brief, with a projected burn-off by 0900. The surface wind was from the northwest at eight knots, gusting twelve, which was nothing. The inversion layer that had been causing low-level turbulence through the earlier part of the week had dissipated. It was, by the measurement of the things that mattered to a man who was going to fly fast and close to the ground for several hours starting at 0730, an excellent morning.

He thought about the missiles.

This was what Bhatia did, in the way that other men might review cricket scores or think about the previous evening's dinner. He thought about weapons with the specific sustained concentration of someone who understood that a weapon was not hardware but a tactical problem — that the value of any airborne weapon was not what it could do in isolation but what it permitted a pilot to do differently than he could do without it, and that the question of what it permitted was a more interesting and more difficult question than the question of what it was made of.

He had flown with the Astra BVR missile twice in trials. He had used the Kaumodaki in live fire against ship targets in the evaluation series off the Saurashtra coast in April. He understood those weapons. They were in doctrine. They were in the muscle memory of the IAF's most capable squadrons in the way that things entered muscle memory, which was through repetition until the weapon's envelope was as familiar as the aircraft's own flight characteristics.

The three weapons he was here to evaluate today were not in doctrine. They were in their final qualification series before doctrine, which was a different condition and one that placed a specific kind of responsibility on the men doing the evaluation. What he assessed, what he wrote, what he said in the debrief this evening — these things would shape how the IAF used these weapons for the next fifteen years, and shaping how the IAF used weapons for fifteen years was not something Bhatia took lightly even on mornings that offered excellent flying weather.

He finished his tea and went inside to the briefing room.

The briefing room at the Pokhran forward operating base was not a comfortable room. It had been built for function rather than for the comfort of the people who used it, and its function was to contain, simultaneously, the tactical display boards, the weapons data terminals, the secure telephone lines, and the twelve to fifteen people who needed to be in the same space discussing things that could not be discussed in any other space. It had a whiteboard along one wall and a large-scale chart of the range along the opposite wall and a row of grey metal folding chairs that had been manufactured with an apparently deliberate indifference to the human lumbar spine.

Present, when Bhatia entered, were the eleven other members of the TACDE evaluation team, the two representatives from Shergill Aviation's weapons division who had flown in from Bangalore the previous evening, and Group Captain Yusuf Hamid Faruqi, who commanded the evaluation series and who was standing at the front of the room with his hands behind his back looking at the range chart in the particular way that commanding officers look at things they already understand completely but look at anyway because the looking communicates authority more efficiently than anything else they could do with their hands.

Faruqi was forty-four, a Hyderabadi Muslim who had entered the IAF in 1952 and had spent twenty-four years acquiring the specific kind of credibility that exists only at the intersection of exceptional flying ability and exceptional analytical intelligence. He had been shot at over the western front in 1965 and had survived it, which he considered a matter of timing and positioning rather than good fortune, and had come home from that war with a specific fury about the quality of weapons his aircraft had been equipped with relative to the quality of pilots flying them. The pilots were excellent. The weapons were a generation behind. He had spent the eleven years since making certain that when the next evaluation series arrived, those two categories were in better alignment.

He had been waiting for these particular weapons for three years.

"Sit down," he said, without turning from the range chart.

Everyone sat down.

"Today's programme," Faruqi said, turning now. "Three weapons, three test series. Shefali close-range air-to-air missile, first and second series, 0730 to 1000. Nishith air-to-ground strike missile, first series, 1030 to 1230. Sudarshan radar-killing missile, full qualification trial, 1400 to 1600. Recovery and debrief from 1700." He looked at the room. "We will not be rushed. If a test series requires a repeat run due to data quality, we repeat. If a weapon performs in a manner that requires a second interpretation of the data before we draw a conclusion, we wait for the second interpretation. This evaluation is the final qualification series before these weapons enter service doctrine. The weapons will be in service for a long time. The evaluation will happen once. We take the time we need."

He nodded to the Shergill Aviation representatives.

The senior representative was a woman named Dr. Saraswati Venkat, forty-one, from Coimbatore, who had joined Shergill's weapons division in 1972 following a doctoral programme in aerospace engineering at IIT Madras and who had led the propulsion integration team on the Shefali and Nishith since the programme's second phase. She was compact, economical in her movement, and had the specific quality of engineers who have been explaining complex systems to non-engineers for many years, which is a quality of absolute clarity without condescension. She stood at the front of the room and opened a large technical folder and began.

"I'll begin with Shefali," she said, "because she is the most straightforward of the three in terms of what today's series is evaluating, and because I want the pilots who will fly the intercept profiles to understand specifically what we're testing rather than simply flying the profile and leaving the interpretation to us." She set a large-format diagram on the easel — a cutaway technical drawing of the Shefali missile in profile, with its major systems labelled in clean military-standard annotation. "What you're looking at is a short-range air-to-air missile — a weapon designed to kill enemy aircraft at close range, in the kind of turning, close-quarters fight that happens when two jets are within a few kilometres of each other. What makes Shefali different from every other close-range missile in any air force today, I'll explain now."

She paused to let the room settle.

"Every close-range air-to-air missile currently in service anywhere in the world — the American Sidewinder, the Soviet Atoll, the French Matra Magic — works on the same basic principle. The missile has a heat-seeking sensor in its nose that looks for the heat coming from an enemy aircraft's jet engine exhaust. To get a lock, the launching pilot has to fly behind the enemy aircraft, pointing his nose roughly at the enemy's tail, so the sensor can see the hot exhaust. If the enemy aircraft is flying across you, or toward you, the sensor cannot see the exhaust and cannot lock. The enemy knows this. His defensive tactic is simple: turn hard, make himself a crossing or head-on target, and the missile cannot chase him." She paused. "The pilot who cannot get his nose behind the enemy is a pilot who cannot fire. Getting behind the enemy in a turning fight is the entire problem of close-range air combat. Everything IAF pilots train for in close combat is essentially: how do I get behind him before he gets behind me."

The room was very still.

"Shefali eliminates that requirement," Dr. Venkat said. "The Shefali does not chase exhaust. Its sensor — a cryogenically cooled array, meaning it's chilled to extreme cold so it can detect even the faint heat signature of an aircraft's fuselage and canopy, not just its engine exhaust — can lock onto an enemy aircraft from any direction. Front, side, above, below. It does not need the tail. And it is coupled to a helmet-mounted sight, which means the pilot points the missile not with his aircraft's nose but with his eyes. Wherever the pilot looks, the missile's sensor follows. The pilot sees the enemy, he looks at him, the missile locks, he fires."

Squadron Leader Pradeep Nair, sitting beside Bhatia, said: "You're saying the pilot can fire at an aircraft that isn't in front of him."

"The pilot can be flying north," Dr. Venkat said, "see an enemy aircraft to his left, turn his head to look at it, and fire. The missile will fly left while his aircraft continues north. The firing pilot does not even need to point his aircraft at the enemy at all. He needs only to look." She paused. "The sensor can track a target up to ninety degrees away from the direction the aircraft's nose is pointed — ninety degrees, which is a right angle, which means directly to the side. Think of a clock face. The aircraft's nose is pointing at twelve o'clock. Shefali can be fired at an aircraft anywhere from nine o'clock to three o'clock, across the entire front half of the sky, without the pilot needing to turn his aircraft at all. In the turning fight, when both aircraft are pulling hard and the geometry is changing every second, this changes everything."

Bhatia said, from the third row: "Walk me through the minimum range. Three hundred metres you said."

"Three hundred metres," Dr. Venkat confirmed. "The minimum range for the Sidewinder is approximately nine hundred metres — inside that, the missile has insufficient time to arm and guide. Shefali can be fired from three hundred metres and can guide and kill from that distance. In practical terms this means Shefali can be fired when the two aircraft are almost on top of each other — inside what pilots call the merge, when both jets have already crossed and are within gun range of each other. No current air-to-air missile in any inventory can operate in that geometry. Pilots who reach the merge currently have only guns. Shefali gives them a missile option inside the merge."

"Range at the other end," Faruqi said.

"Twelve kilometres at altitude. At altitude and in a tail-chase geometry — the traditional firing position — it can kill at twelve kilometres. But the designers did not optimise it for long range. Shefali was built for the close fight. Its purpose is to give the Indian pilot a decisive advantage in the turning, close-range engagement where currently both sides are roughly equal because both sides are limited to the same tail-chase geometry."

She turned to the next diagram. "The warhead is a directional type — when it detonates, rather than throwing fragments in all directions equally, it throws a concentrated pattern of tungsten metal cubes in the direction of the target. Combined with a laser proximity fuse that detonates the warhead at the optimum moment rather than requiring a direct hit, the missile can kill an aircraft even if it misses by several metres. The directional pattern means the tungsten reaches the target's most vulnerable areas with far higher density than an omnidirectional warhead would achieve."

Faruqi spoke from his position at the side of the room. "What's today's series evaluating specifically?"

"Two things," Dr. Venkat said. "First: we need to verify that the helmet-mounted sight works correctly when the pilot is under high G-loading in a turning fight. G-force in a hard turn — six, seven times the force of gravity — affects everything. The pilot's head is heavy, his neck is straining, his vision is affected. The question is whether the sight's tracking of where the pilot's eyes are pointed remains accurate enough under those conditions to give the missile a valid lock. We have tested this on the ground and in gentle manoeuvres. Today we test it in a real turning fight under real G."

"Second: we will fly a scenario where the launch pilot fires Shefali at a target drone in a crossing geometry — not chasing it, crossing it — at ranges between three hundred and eight hundred metres. This is a shot that no current weapon can take. We need to prove Shefali can."

"The target drone," Bhatia said.

"Mirach target drone, flying a hard sixty-degree banked turn at 450 knots. It will be presenting its side to the launching aircraft, not its tail. Any other missile in the IAF's inventory sees that drone and cannot engage. Shefali sees it and fires."

Bhatia was looking at the ceiling, which was what Bhatia did when he was turning a tactical problem over from every angle. The pilots who knew him recognised it.

"The doctrine implication," he said slowly, "is complete. If Shefali fires from any angle within that front-hemisphere coverage, the optimal tactic is no longer to try to get behind the enemy. The optimal tactic is to enter the fight with high speed and energy, create a moment where the enemy aircraft is anywhere in that front ninety-degree zone, and fire before he can deny you that geometry." He paused. "The entire close-combat doctrine the IAF has trained for twenty years is built around getting the tail shot. This missile makes the tail shot optional."

"Which is one of the things we're here to develop doctrine for," Faruqi said. "The missile is a new capability. The doctrine doesn't exist yet. We're going to develop it."

"We're going to start developing it today," Bhatia said.

"We're going to start developing it today," Faruqi agreed. "The full doctrine takes longer than one day. But yes. We start today."

The Shefali firing series began at 0742, eighteen minutes later than scheduled because one of the Mirach drones had required a last-minute transponder check that took seventeen minutes. This delay caused the test director, a Squadron Leader named Arun Ghosh from the TACDE's weapons analysis cell, to make a note in his test log with the economical fury of someone who considers seventeen minutes a personal affront to the organisation of time.

Bhatia flew the first intercept in a TACDE S-27 Pinaka.

The S-27 was the right aircraft for this work in ways that went beyond the obvious. It was Shergill Aviation's frontline air superiority fighter, equipped with the Kaveri Mk1 engine that gave it more raw power than any equivalent aircraft in the region, and the TACDE's evaluation fleet of eight S-27s had been modified with the Shefali launch rail on the port inner wing station and the helmet-mounted sight integration wired into the aircraft's weapons management system. The cockpit was wide enough, and the canopy tall enough, that the helmet-mounted sight — a unit that added meaningful weight to a standard flying helmet — did not press against the canopy frame when the pilot turned his head to the limits the system required. This detail had mattered in the ground testing. It had been the reason TACDE's previous evaluation aircraft, the older MiG-21 fleet, had been specifically excluded from the Shefali programme: the MiG-21's narrow cockpit and low canopy could not accommodate the HMS without restricting the pilot's head movement to the point where the weapon's angle coverage became meaningless. You could not look ninety degrees sideways in an aircraft designed for a man sitting in a bath. The S-27's generous cockpit geometry was not incidental to the Shefali programme. It was a prerequisite.

The HMS was not the most elegant piece of equipment Bhatia had ever put on his head. It was a modified flight helmet with an optical tracking system mounted at the left temple that read the helmet's orientation against a set of reference emitters fixed inside the cockpit, translating the position and direction his head was pointing into a continuous signal that the Shefali's sensor could receive. The result was that wherever Bhatia looked, the missile knew. Its sensor slewed to follow his gaze.

The conceptual simplicity of this was, Bhatia had concluded after his first ground test of the system two months ago, slightly misleading. The execution was not simple. The challenge was not in the slewing — the sensor moved fast and accurately — but in the human factor. A pilot in a hard turn does not look at a target with the stable, deliberate gaze of a man reading a sign. He looks with the rapidly shifting, high-acceleration gaze of a man who is simultaneously managing his aircraft, his speed, his position in space, his communications, and the specific physical strain of six or seven times the force of gravity on his body and his head. Under that G-load, the pilot's head weighs as much as a sack of bricks. The muscles of his neck are fighting to hold it still enough for the tracking system to read where he's looking. The question the first trial needed to answer was whether the helmet's tracking was accurate enough under those conditions to give the missile a valid lock.

He took the intercept profile from the range controller at 15,000 feet over the northern range area, the Mirach drone already up and running its profile at 12,000 feet below him. The drone was flying a sustained sixty-degree banked turn at 450 knots — a tight, fast circle, simulating an enemy fighter in a hard defensive break. From Bhatia's position, he could see its geometry clearly: the drone was presenting its left side to him. A crossing shot. A target presenting its side and not its tail. A target that any other missile in the IAF's inventory would require him to pursue and get behind before engaging.

He armed the Shefali station.

The tone in his earpiece changed. Not the sharp howl that the Sidewinder produced when it smelled hot exhaust — this was a steady, confident note, lower, more assured, the signal of a sensor that had found the entire aircraft and not just the heat of its engine. It had the quality of something that knew what it was looking at.

He looked left toward the drone. The drone was well to his left, maybe sixty-five degrees away from the direction his aircraft was pointed. The missile's tone held steady. It was tracking what his eyes were tracking.

He set up the crossing geometry the test required: closing at 480 knots toward a point that would take him across the drone's flight path at approximately 400 metres, both aircraft presenting their sides to each other at the moment of maximum proximity.

In any other configuration — any other missile in the IAF's current arsenal — what he was about to do was not a missile shot. It was a guns pass. A guns pass at 900 knots combined closure, crossing geometry, 400 metres range, required precise gunnery against a turning target, and even in excellent conditions it was a shot that trained pilots missed.

He looked at the drone.

The seeker tone shifted fractionally — a slight rise in pitch, steady, not pulsing. Acquisition confirmed.

He fired.

The Shefali came off the rail with a sound unlike anything else he had fired from an aircraft — not the thunderclap of a gun, not the long, building roar of a heavy air-to-ground weapon. A sharp crack of motor ignition followed immediately by the hiss of something moving very fast, and then it was gone, already turning in mid-air before it had cleared the aircraft's wing shadow, the thrust-vectoring fins at its tail bending its path toward the drone in a curve that looked from the cockpit — in the fraction of a second he had to see it before his own velocity carried him past the geometry — like something that had already decided where it was going and was simply going there.

The drone ceased to exist as a functional drone approximately 0.7 seconds after the Shefali reached its proximity fuse decision point.

The missile had not hit the drone directly. It had detonated 1.8 metres from the drone's fuselage centre line, precisely at the wing root junction where the structure was most vulnerable, and the directional fragmentation pattern — the concentrated spray of tungsten cubes aimed like a shotgun at the kill zone rather than scattered in all directions — had opened the drone's wing attachment and sent it into an uncontrolled tumble from which no aircraft returns.

Bhatia said, on the range radio: "Good kill."

Range control: "Confirmed. Telemetry nominal. Splash One."

In the control van on the range road, Arun Ghosh wrote in his test log: 0751:23. Shefali Test 1-A. Angle from aircraft nose to target at launch: 67 degrees left. Range at launch: 412 metres. Detonation: confirmed. Target kill: confirmed. Pilot: Wg Cdr Bhatia. Notes: Controlled crossing geometry, not the high-G tracking test — that comes in the turning fight series.

The turning fight series was the second set of Shefali trials, beginning at 0830 after a twenty-minute recovery and debrief on the first series data.

Bhatia flew it with Squadron Leader Pradeep Nair as the adversary aircraft. Both men flew S-27 Pinakas. The scenario was constructed to replicate as accurately as a managed test could replicate the conditions of a real close-combat turning engagement: two aircraft approaching each other at high speed from opposite directions — what fighter pilots called the merge — and then the fight that happened after the merge, when both aircraft were trying to manoeuvre into a firing position against the other.

In a conventional close-combat engagement — with any missile currently in any air force's service — the fight after the merge has a single objective for each pilot: get behind the other aircraft. Get your nose pointed at his tail. From behind, you can see his engine exhaust, your missile can lock, you can fire. The pilot who achieves that position wins. The pilot who cannot prevent the other from achieving it loses. The entire art of close air combat, the reason IAF pilots spend hundreds of hours training turning and rolling and managing their speed through a fight, is the pursuit of that one geometry: nose behind tail.

Shefali changed what the fight was for.

Bhatia was not trying to get behind Nair. He was trying to achieve any moment — any fraction of a second — in which Nair's aircraft was somewhere in the front ninety-degree zone of his own aircraft's coverage. Not behind Nair. Simply not in a head-on collision course. Anywhere in that huge front quarter of the sky, and Bhatia could fire.

Nair understood this. He had been in the same briefing room and was not a man who failed to process tactical information. His defensive response was logical: he was trying to deny even that geometry. If a side-on crossing shot was now lethal, he could not afford a crossing shot. He had to stay either directly head-on to Bhatia — in which case neither could fire safely — or directly running away, tail-on, which was also a firing geometry but one that Bhatia had to chase to achieve.

The result was the most unusual defensive flying Nair had done in his career. He was trying to fly at Bhatia or away from Bhatia at all times, never across him, a constraint that turned the normal tactics of a turning fight inside out. A turn that would normally be useful for energy management or position was now potentially suicidal because it briefly presented his aircraft's side. He flew in tight, committed reversals — turning hard toward Bhatia, then immediately reversing away — sacrificing the smooth, energy-efficient defensive turns that training had drilled into him.

At the forty-seventh second of the engagement, Nair's defensive reversals had forced him into a situation where his turn radius was smaller than Bhatia's and his energy state was lower — he had been spending speed to make those tight reversals. Bhatia was higher, faster, with more energy in reserve.

In a conventional fight, with a conventional missile, Bhatia would now need to chase Nair's tail and wait for Nair to make a mistake. It might take another minute. It might not happen.

Bhatia unloaded his G — stopped the turn, briefly flew straight — for two seconds. He pulled the S-27's nose twenty degrees above the horizon. He rolled inverted.

From inverted, looking down through the top of the canopy at the earth below — and at Nair's aircraft passing below him in a crossing geometry — he was looking roughly eighty degrees away from the direction his aircraft's nose was pointed. Under normal circumstances, that angle was militarily meaningless. No missile could use it.

The Shefali tone was steady in his earpiece. The sensor was following his eyes. It had Nair's aircraft.

He fired.

The Shefali left the rail while Bhatia's aircraft was inverted, nose high, eight G loading from the manoeuvre, the pilot looking nearly sideways through his canopy at a target below him. The missile turned in mid-air the moment it left the rail, vectoring immediately toward Nair's aircraft on the geometry that Bhatia's gaze had given it.

Nair received the simulated kill tone in his earpiece — the range control's systems confirmed the weapons solution and transmitted it to the adversary aircraft — and said on the frequency: "You're kidding."

"Test data will confirm," Bhatia said.

"I know it will. That's not what I mean." A pause. "I was in a defensive turn, specifically designed to deny a crossing shot. You pulled inverted, fired while upside down with your nose not pointed anywhere near me, and the missile still found me. There is no defensive manoeuvre I have been taught that accounts for that."

"That's why we write new doctrine," Bhatia said.

In the control van, Arun Ghosh was writing very fast and had run out of space in one column of his test log and was continuing in the margin.

The Nishith series began at 1035, after a forty-minute break in which the test team ate what the range's mess could provide and Bhatia spent most of the time at the range chart with Faruqi, working through what the morning's Shefali results meant.

"The defensive problem the Shefali creates," Faruqi said, "is that the pilot defending against it cannot use the standard defensive break — the hard turning manoeuvre that currently gives you a crossing aspect and denies the tail shot. Against Shefali, a crossing aspect is a firing solution. Nair found the only real defence is to deny any side-on geometry entirely."

"Which means the defending pilot has to fly directly at you or directly away from you," Bhatia said. "He cannot turn across you. Cannot break across you. Any turn that presents his side is a kill geometry." He looked at the range chart. "What this does to the fight is force the defending pilot into very predictable flight paths — toward you or away from you. Both of those are easier to manage than the unpredictable turning defence we currently have to handle."

"He becomes easier to kill," Faruqi said.

"The defensive tactic against Shefali makes the defending pilot easier to kill, yes. The correct defence is not a manoeuvre. The correct defence against a pilot with Shefali is to never let him get close enough in the first place." Bhatia paused. "Which means the missile changes the engagement geometry even before the merge. The enemy pilot who knows his adversary carries Shefali will try to convert the fight to a long-range engagement where he never has to deal with this problem. Which means the long-range missile — Astra — becomes the weapon that forces the enemy into the close fight he doesn't want. And Shefali is what kills him in that fight if he can't avoid it."

"The two missiles work together," Faruqi said.

"They were designed to work together," Bhatia said. "The short-range capability makes the long-range capability more threatening, and vice versa. An enemy who avoids long range gets Shefali. An enemy who tries to close and use his own short-range weapon gets Astra. There is no clean option."

"Doctrine development series in six weeks," Faruqi said. "Full combined-arms scenario with both weapons, against all adversary profiles we can construct."

"Agreed." Bhatia looked at his watch. "Nishith."

They went back to the briefing room.

The Nishith was a different kind of weapon entirely.

Where Shefali was a missile designed for the personal, intimate geometry of two aircraft trying to kill each other in the same piece of sky at close range, Nishith was a missile designed for something entirely different: letting a pilot kill something on the ground without having to fly close enough to the ground defences to be killed himself.

Dr. Venkat presented the Nishith data at the same easel, with a different diagram.

"The Nishith is an air-to-ground strike missile," she said. "Before I give you the specifications, let me explain the problem it solves, because the problem is one that has been killing IAF pilots since 1965.

"When a pilot is ordered to destroy a target on the ground — a radar installation, a bridge, a fuel depot, an enemy command bunker — he has to get close enough to hit it. With unguided bombs, which is what the IAF has been dropping for thirty years, close enough means the pilot must overfly the target or come within one to three kilometres of it. At that range, every anti-aircraft gun, every surface-to-air missile, every radar-guided weapon system defending the target can engage him. The attack run — the few seconds when the pilot is close enough to hit the target — is also the few seconds when the defences have the best shot at him. This is not a problem of pilot skill. The best pilot in the world cannot fly through a defended zone without accepting the risk of being killed. It is a problem of physics: you cannot hit what you cannot reach, and reaching it means being close to what defends it."

Squadron Leader Tejinder Walia, who had been flying strike missions for a decade, was looking at the diagram with an expression that Bhatia, from two seats away, recognised as a man who has spent ten years accepting a fundamental constraint of his profession and is now being told the constraint no longer applies.

"The Nishith solves this by separating the launching aircraft from the point of impact," Dr. Venkat continued. "The pilot does not fly to the target. He flies to within launch range — and then the missile flies the rest of the way. The pilot launches from safety, turns, and goes home. The missile does the dangerous part." She pointed to the diagram. "Range from launch to impact: twenty-eight kilometres at low altitude and high speed, extending to forty-two kilometres from medium altitude. The SA-2 surface-to-air missile system — which is the primary threat to IAF strike aircraft against any modern air defence network — has an engagement range of approximately forty-five kilometres against high-altitude targets but is largely ineffective at very low altitude and has limited coverage between eight and fifteen kilometres altitude. The Nishith launch envelope is specifically designed to keep the launching aircraft outside the SA-2's most effective engagement zone. The missile flies through the threat. The aircraft does not."

"How does the missile know where the target is from twenty-eight kilometres?" Walia asked.

"Two guidance stages," Dr. Venkat said. "In the first stage — the long approach — the missile uses inertial navigation. Think of it as a very precise internal compass and timer: at the moment the pilot fires, the aircraft's navigation system passes the exact coordinates of the target to the missile's computer, and the missile flies to those coordinates using its own internal sense of direction and distance, with no further input needed from the pilot or from any external signal. This gets the missile to within roughly one kilometre of the target."

"That's not precise enough to hit most tactical targets," Walia said.

"Correct. Which is why there is a second stage. Once the missile is approximately six kilometres from the target, a laser system takes over. Either the launch aircraft or a second aircraft in the flight illuminates the target with a laser beam — a narrow, invisible beam of light aimed at the specific spot to be destroyed. The missile's nose sensor detects the reflected laser energy bouncing off the target and homes in on it with precision. The laser designation brings the impact point to within five to twelve metres of the aiming point. Against most tactical targets — a radar van, a bridge span, a hardened bunker entrance — five to twelve metres is a kill."

"The laser-equipped aircraft has to get close to designate," Walia said.

"Eight kilometres is the effective designation range of the SI-75 designation pod. At eight kilometres altitude and medium height, the designating aircraft is outside the engagement envelope of all SA-2 and most SA-3 radar-guided missile systems. It is not inside the threat. It is outside the threat, painting the target with a laser while the missile does what no pilot can safely do — fly through the defended zone to the impact point."

"Warhead options," Faruqi said.

"Three, depending on the target type. A 250-kilogram blast-fragmentation warhead for soft targets and open emplacements — vehicles, fuel storage, radar dishes, command tents. A 120-kilogram hardened penetrator for reinforced concrete structures — command bunkers, underground facilities, radar control buildings built to resist air attack. This penetrator is designed to punch through two metres of reinforced concrete before detonating inside the structure, which is how you destroy what is inside rather than simply damaging the exterior. And a 180-kilogram tandem HEAT warhead for armoured targets — a warhead specifically engineered to defeat the explosive reactive armour that modern tanks are beginning to carry. The arming crew selects the warhead for the mission before flight. The aircraft's weapons computer reads what is installed and adjusts the launch calculations accordingly."

Walia had been making notes with increasing speed. He stopped, set his pen down, and looked at the diagram in silence for a moment.

"In 1965," he said, finally, "the IAF lost aircraft on strike missions against defended targets at rates that mission planners described as the cost of the mission. The cost of the mission meant pilots. It meant people dying because the only way to hit the target was to fly through what was defending it." He picked his pen up again. "This missile says the cost of the mission is the missile. Not the pilot."

"That is the design intent," Dr. Venkat said.

"Then what we're testing today," Walia said, "is whether the design matches the intent."

"Yes," she said. "That is exactly what today is for."

The Nishith live fire began at 1052.

The evaluation was designed in three phases. Phase one: pure inertial accuracy — two Nishith rounds fired against a fixed target at thirty kilometres, no laser guidance, inertial navigation only, to confirm the missile's internal navigation was performing to specification. Phase two: laser-guided accuracy against a moving target — a radio-controlled jeep, driven in an unpredictable direction-changing pattern across the range floor to simulate a mobile radar vehicle trying to reduce its vulnerability. Phase three: penetrator warhead against a concrete bunker representative built at the south end of the range.

Walia flew Phase One.

He took his S-35 Tejas-M — the single-engine multi-role aircraft that Shergill Aviation had designed for exactly this kind of low-level, high-speed strike work — in at 500 feet, 0.83 Mach, terrain-following across the Thar's surface at the kind of altitude where the desert was not a landscape but a texture passing beneath him at most of the speed of sound. His weapons systems officer in the rear seat was managing the navigation computer, confirming the target coordinates that had been loaded before takeoff, monitoring the countdown to launch parameters.

At thirty-one kilometres from the target, the navigation computer confirmed readiness.

At thirty kilometres, Walia fired.

The Nishith came off the right wing station and immediately pitched upward — climbing away from the low-altitude launch into a steep arc, trading the aircraft's speed for altitude to achieve the energy state the long-range flight required. The missile's computer was calculating its trajectory in real time, flying a curved path that would eventually bring it back down toward the target coordinates from above, the guidance system threading it through a path optimised for both range and accuracy.

Walia did not follow the missile. He pulled up, turned north, began recovering to medium altitude. The missile was gone. It would arrive at the target coordinates or it would not. There was nothing more he could do about it.

The target was a concrete marker block, three metres by two metres, standing alone on the range floor at exactly thirty kilometres from the launch point.

The Nishith impacted 8.3 metres from the block's centre.

In the control van, Ghosh noted it: Nishith Test 1-A. Inertial guidance only. Range 30km. Impact distance from target: 8.3 metres. Specification: 15 metres maximum. PASS.

The second round, fired ten minutes later by Flight Lieutenant Gurpreet Dhaliwal in his S-27 Pinaka modified for the weapons trial, impacted 6.1 metres from the same target.

Nishith Test 1-B. CEP: 6.1 metres. PASS.

Phase Two was the test that mattered most for the doctrine writers.

The moving target — a radio-controlled jeep fitted with an infrared targeting marker that the designation laser could find — was moving at approximately fifteen kilometres per hour in a random direction-changing pattern across the range floor. The tactical scenario it replicated was the Nishith employed against a mobile SA-6 radar vehicle — a radar system that Pakistani and Egyptian forces had used effectively in 1973, that moved between firing positions to avoid being targeted, and that had proven extremely difficult to destroy precisely because its mobility meant it was never where your attack aircraft expected it to be.

Wing Commander Harish Chandra Pandey flew the designation aircraft, a two-seat S-35 Tejas-M trainer variant configured with the SI-75 designation pod on the centreline station. His weapons systems officer in the rear seat, Flight Lieutenant Suresh Balakrishnan, would operate the pod — tracking the moving jeep with the laser spot throughout the terminal phase of the Nishith's approach.

The designation job was harder than it sounded. The aircraft was in a banked orbit at eight kilometres from the target, the pod's stabilised optics keeping the laser spot on the moving vehicle. The pod had to compensate for the aircraft's own movement, the vehicle's movement, and maintain stable tracking throughout the twelve or so seconds of the Nishith's terminal homing phase. Pandey and Balakrishnan had been training with the SI-75 for four months. The coordination had become a kind of dialogue between pilot and operator — Pandey adjusting his orbit to keep the pod's gimbal geometry optimal, Balakrishnan calling speed and bearing changes on the target, both men making continuous small corrections to a moving problem with no clean solution.

Walia fired the Nishith from thirty kilometres, medium altitude.

Pandey established his designation orbit at eight kilometres. Well outside the SA-2 envelope. Well outside the SA-6's own radar detection range for low-signature aircraft.

"Laser on," Balakrishnan said.

Range control confirmed the spot was on target.

The Nishith's terminal phase began at six kilometres. The missile's sensor switched from inertial coast to laser homing, reading the reflected energy from Balakrishnan's spot on the moving jeep. The guidance computer began its intercept calculation — not where the jeep was, but where the jeep would be when the missile arrived, the classic lead-angle calculation that every gun-armed pilot knew but was here being performed by a computer in real time at much greater precision and speed than any human could manage.

The jeep was hit at its left front quarter. The vehicle was destroyed by the blast-fragmentation warhead's pattern. The impact point was within the effective kill radius of the strike.

Range control, on the radio: "Direct hit on moving vehicle. Good kill."

In the designation aircraft, Balakrishnan said, quietly: "Yes."

Pandey said nothing. He brought the aircraft around and looked at the smoke rising from the range floor and thought about mobile radar vehicles and about the fact that an IAF pilot had just killed a moving target from twenty-two kilometres of horizontal separation, never closer than eight kilometres to the target himself, using a weapon that navigated itself with a precision no pilot's hands could match.

He keyed his radio. "TACDE Control, Designation One. Phase Two complete. Laser guidance positive. Target mobility profile replicated. Request Phase Three clearance."

"TACDE Control. Phase Three is clear."

The Phase Three target was a concrete structure built at the south end of the range to the specifications of a hardened command bunker — the kind of installation that an adversary's radar control centres and underground command nodes were built to. Two metres of reinforced concrete, poured over three weeks by the range construction team. It had cost twelve thousand rupees to build. It would cost one Nishith penetrator round to make it irrelevant.

Flight Lieutenant Kulwant Singh Randhawa flew the Phase Three shot in a single-seat S-35 Tejas-M at 8,000 feet, Mach 0.89. He was twenty-seven years old and the youngest pilot in the evaluation series by six years. He had been in TACDE for fourteen months and had been selected for this evaluation because Group Captain Faruqi had a specific theory about who should fire each weapon in its qualification series: the weapon that the IAF would use for the hardest targets, against the most dangerous defended positions, should be evaluated by the youngest pilot, because the youngest pilot was the one who would still be flying it when those engagements happened for real, and whose understanding of what the weapon could do needed to be first-hand rather than inherited from a briefing.

Randhawa flew the delivery profile with the economy of movement that comes from thorough preparation — no wasted control inputs, no hesitation in the sequence. The penetrator warhead required the Nishith to impact at a specific angle — between fifty and seventy degrees from horizontal — to achieve clean penetration rather than deflecting off the concrete surface. The guidance system had a terminal phase correction mode specifically for penetrator deliveries, adjusting the final approach trajectory to hit within that window.

Pandey's aircraft designated from eight kilometres. Balakrishnan held the laser spot steady on the bunker's roof — a stationary target, no direction changes, the variables reduced to just the aircraft's own movement and the need for precise spot placement.

The Nishith impacted the bunker's roof at sixty-three degrees from horizontal. Within specification. At the exact point of the laser spot.

The penetrator warhead punched through two metres of reinforced concrete and detonated inside the structure. The bunker's roof failed. The internal pressure wave destroyed the simulated contents.

The range safety officer, 1.2 kilometres back: "Detonation confirmed, subsurface. Structure breach confirmed. Target rendered non-functional."

Ghosh's log: Nishith Test 3-A. Penetrator warhead. Impact angle: 63 degrees — within specification. Penetration: confirmed. Subsurface detonation: confirmed. Structure assessment: destroyed. PASS. All Phase 1-2-3 objectives met. Nishith qualification trial: recommend PASS for service entry.

He looked at what he had written. Added one line, below the box: Tactical significance — a pilot can now kill a hardened command bunker from 32km away without entering the defended zone. IAF strike doctrine requires fundamental revision.

The midday break was forty minutes. Faruqi extended it by twenty minutes without announcement, simply by not ending it, and the team ate lunch at the slow pace of men who had much to think about.

Walia was quiet through most of it. He ate carefully and looked at his notes and looked at the range chart on the wall and did not say anything for fifteen minutes.

Then he said, to no one in particular: "In 1965, I flew a Hawker Hunter against defended targets on the western front. The planning assumption was that we would lose some aircraft on every strike package. Not might lose. Would lose. The number was built into the mission plan. Acceptable losses per sortie. It was considered a fact of the work." He paused. "The Nishith says that assumption is wrong. Not optimistically wrong — factually wrong. The aircraft does not have to enter the envelope that produces those losses. The missile enters the envelope. The aircraft does not."

Nobody at the table said anything. There was nothing useful to add to it.

The Sudarshan evaluation began at 1412, slightly behind schedule. The midday conversation had run longer than the timeline permitted, which Faruqi had allowed because some conversations were worth the schedule slippage.

The Sudarshan was the weapon that none of the team had a direct frame of reference for, because the capability it represented — an anti-radiation missile, a missile specifically designed to hunt and kill enemy radar systems — was not something any IAF pilot had previously operated. It was a new category.

Dr. Venkat presented it with particular care.

"The Sudarshan," she said, "is designed for one mission: find and destroy enemy radar systems. I want to explain why this matters as a mission before explaining how the missile achieves it.

"Every modern air defence system — every surface-to-air missile system that threatens IAF aircraft — depends on radar. The radar searches the sky, finds the aircraft, tracks it, guides the missile to it. Without the radar, the system is blind. A blind SAM system cannot fire at what it cannot see. If the radar dies, the entire air defence battery around it — the missiles, the guns, the command vehicles — becomes inert. You do not have to destroy every component of an air defence system. You only have to destroy the radar at its centre. The rest becomes expensive scrap metal with nobody who can aim it."

She looked at the pilots.

"The problem is that radars emit radio signals. To do their job, they have to transmit. And a transmitted signal is a beacon. It says: here I am, here is my exact position. The Sudarshan is a missile that listens for that beacon, locks onto it, and flies to its source. A radar that is transmitting is advertising its own location to the Sudarshan. The more it transmits, the more precisely the Sudarshan knows where it is."

"What if the radar operators turn it off?" Faruqi said. "The standard counter-tactic."

"They have two options," Dr. Venkat said. "They can leave the radar on and be killed by the Sudarshan. Or they can turn the radar off, in which case their own SAM system goes blind and cannot engage the aircraft attacking them." She paused. "Either way, the radar is no longer guiding missiles at IAF pilots. The Sudarshan creates a situation where the radar operator must choose between being killed and being useless. There is no option that allows him to both survive and do his job effectively."

"That," Nair said, from the second row, "is an elegant problem to impose on someone."

"The engineers thought so," Dr. Venkat said. "The Sudarshan's seeker covers the radio frequency range from one gigahertz to eighteen gigahertz. That covers every radar system currently deployed in any air defence network we are aware of — SA-2, SA-3, SA-6, Hawk, Bloodhound, all of them. The pilot does not need to know in advance what type of radar is at the target. The Sudarshan finds any radar in that frequency band, categorises it by its signal characteristics, and homes on it."

"What if the radar shifts frequency?" Banerjee's question — not a pilot but the Shergill Aviation propulsion engineer who had joined the afternoon session. He was quickly silenced by a look from Dr. Venkat's colleague.

"The range today includes a frequency-shifting test," Dr. Venkat said. "That is specifically what the third shot is evaluating. We will address it when we get there."

She continued: "Launch range. Minimum range: five kilometres. Maximum range: forty-five kilometres at low altitude, sixty-two kilometres at medium altitude. At sixty-two kilometres the launching aircraft is well outside the detection range of most of the radar systems it is targeting. The SA-2's radar — the Fan Song — can detect a fighter-sized target at approximately 60 kilometres in ideal conditions. At 62 kilometres, launching the Sudarshan, the IAF aircraft is at the edge of detection or outside it. The Sudarshan reaches the radar before the radar crew can be certain the threat is real."

She set the diagram down.

"The missile also has a memory mode. If the enemy radar operators see the missile coming and turn the radar off, the Sudarshan remembers the last position it calculated for the radar and continues flying to that position on internal guidance. The radar can go silent. The missile does not stop. It continues to the last known position of the radar and detonates there. The circular error — the expected miss distance — for a radar that shut down in the first fifteen seconds after launch is approximately six metres from where the radar was when it shut down. Most radar vehicles do not move six metres in fifteen seconds."

"Three shots today," Faruqi said.

"Shot One: against a working search radar transmitting continuously at S-band, fifty kilometres range, inertial and radar-homing combined. Shot Two: memory mode — the radar shuts down at ten seconds after launch. The missile must continue to the last known position. Shot Three: frequency-agile radar — a radar that shifts its transmission frequency three times during the missile's flight, to simulate an adversary radar using electronic counter-measures to try to break the Sudarshan's lock."

"Who's flying Shot Three?" Bhatia asked.

Faruqi looked at him. "You are."

"Good," Bhatia said.

Walia flew Shot One.

The target radar — a working S-band search radar built by the range's electronic warfare section — was transmitting at full power on the south range. From 52 kilometres in his S-35 Tejas-M, Walia could see nothing with his eyes — the radar was invisible at that range. But the Sudarshan's seeker could hear it the moment Walia turned the aircraft toward the range. The signal came in clearly, the seeker began its calculation, and by the time the navigation computer confirmed launch parameters at 50 kilometres, the Sudarshan knew the radar's position to a precision of less than twenty metres.

Walia fired.

The missile climbed off the left outboard station with a slightly different acoustic quality from the Shefali and the Nishith — heavier, more deliberate, a weapon that had more distance to cover and was built for endurance rather than agility. It turned immediately toward the south range. The seeker's antennas along the fuselage sides were continuously reading the radar's signal, continuously refining the position solution with each new measurement.

The radar transmitted continuously throughout the flight.

The Sudarshan impacted eleven metres from the radar antenna assembly. The blast-fragmentation warhead produced a lethal fragment pattern that covered the antenna and its rotation mechanism. A radar that cannot rotate cannot search. A radar that cannot search cannot guide missiles. The antenna was gone.

Ghosh: Sudarshan Test 1-A. Range 50km. Target: continuous transmission. Distance from antenna: 11 metres. Specification: 15 metres maximum. PASS.

Shot Two: Nair flew it. The radar transmitted for nine seconds before its operators simulated shutdown. The Sudarshan, in memory mode, continued to the stored last-known-position.

Impact: 7.4 metres from the antenna's position at the moment of shutdown.

Sudarshan Test 2-A. Memory mode. Shutdown at T+9 seconds. Distance from shutdown position: 7.4 metres. Specification: 10 metres. PASS.

Shot Three.

The test aircraft was an S-27 Pinaka. Bhatia's aircraft.

The target for Shot Three was the hardest problem on the day's programme: a radar that could shift its frequency while the missile was in flight, mimicking the electronic counter-measure capability that Soviet radar designers had been building into new systems specifically to defeat ARM missiles. The range had constructed this test asset at considerable cost because no equivalent existed in the Indian test infrastructure. It could shift its transmission frequency, change its pulse characteristics, and alter its operating band — three times during the Sudarshan's flight, at programmed intervals.

The Sudarshan's seeker had to track through each shift.

Bhatia took his launch parameters at 58 kilometres, 9,500 feet, Mach 0.91. The seeker was reading the target radar's initial X-band transmission clearly. He fired.

At twelve seconds after launch, the radar shifted from X-band to S-band — a completely different radio frequency, a different part of the electromagnetic spectrum. A conventional anti-radiation missile, designed to home on one specific frequency, would lose lock immediately and fly blind.

The Sudarshan's signal-processing software — designed not to track a fixed frequency but to track the radar's entire emission signature, the pattern of how it transmitted rather than the specific frequency — recognised the shift as the same radar source changing channels. The adaptation took 0.4 seconds. The position calculation continued without interruption. The error introduced by 0.4 seconds of adaptation lag was less than two metres at the missile's speed and range.

At twenty-four seconds, the radar shifted to C-band. The adaptation this time took 0.6 seconds — C-band was at the edge of the seeker's frequency range, requiring more processing time for confident re-acquisition. The position error: 3.1 metres.

At thirty-five seconds, the radar shut down entirely.

Memory mode. The Sudarshan's computer had been building its position solution throughout the two frequency shifts, each adaptation adding a small error. Total accumulated positional uncertainty at shutdown: approximately five metres.

The missile flew to the stored position.

It impacted 9.2 metres from the radar antenna assembly.

Specification for memory mode with shutdown in this time window: ten metres maximum.

9.2. A margin of 0.8 metres.

Ghosh stared at his log for a moment before writing. Then: Sudarshan Test 3-A. Frequency-agile target. Two frequency shifts. Target shutdown at T+35 seconds. Distance from antenna: 9.2 metres. Specification: 10 metres. PASS — margin 0.8 metres.

He added below the box, in the manner of a man noting something that needed to be said even in a technical log: This is a pass. It is also a pass that says the C-band adaptation delay needs attention before this weapon is used against a confirmed frequency-agile radar in operational conditions. 0.8 metres is not a comfortable margin.

The debrief began at 1703 and ran for three hours and nine minutes.

Faruqi ran it with the discipline of a man who understood that the debrief was where the day's flying became something the IAF could use for the next twenty years. Each sortie, each shot, each number reviewed in sequence. Dr. Venkat and her colleague present and answering technical questions but speaking only when asked, because the debrief was the pilots' analysis to conduct.

Bhatia presented the Shefali conclusions.

"The helmet-mounted sight tracking is accurate under high G," he said. "The six-and-seven-G manoeuvre data from the turning fight confirms the seeker maintains valid lock during dynamic head movement at those G-loads. There is a measurable increase in tracking lag above seven G — the helmet moves faster than the tracking system can perfectly follow at those loads, producing approximately half a degree of angular error per G above seven. This is within the weapon's guidance correction capacity and does not affect kill probability. It should be understood and documented."

He paused and looked at the room. "The tactical conclusion from today's close-combat trials is simple but its implications are not. The entire doctrine of close air combat has been built around one objective: get behind the enemy. Get your nose on his tail. Every manoeuvre IAF pilots train in the turning fight exists to serve that objective. The Shefali makes that objective optional." He looked at the range chart. "The new objective is: don't be directly in front of him and don't be directly behind him. Be anywhere else. Be to the side. Be above him looking sideways. Be inverted looking through the roof of your canopy. Any angle except the direct collision course is a firing solution."

He looked at Faruqi.

"The doctrine development series must include the defensive doctrine. Any IAF pilot who will carry Shefali needs to know how an enemy pilot would defend against Shefali, because that enemy pilot, if he is well-trained, will eventually figure it out. We figure it out first. We build the counter into the training before the adversary builds the counter into his doctrine."

Faruqi nodded. "Six weeks. Full series. All adversary profiles."

Walia presented the Nishith conclusions standing at the front of the room, which was not standard — the other pilots had presented from their seats. Faruqi let him stand without comment.

"I have been flying strike missions for ten years," Walia said. "I want to describe the Nishith's significance in the terms I understand, which are the terms of what it means to plan and fly a strike into a defended target.

"When I plan a strike mission today, against a defended target, I plan around the assumption that some aircraft in the package will be lost. I calculate approach routes that minimise exposure time inside the defended zone. I plan ingress profiles at very low altitude to avoid radar detection, which means flying at five hundred feet and five hundred knots across unfamiliar terrain, which is where pilots die even before the defences engage. I plan escape routes for aircraft that are hit but may still fly. I plan for aircraft that cannot escape." He paused. "All of this planning accepts that the only way to hit the target is to be in the target's defended zone.

"The Nishith says that planning assumption is wrong. The aircraft does not enter the defended zone. The aircraft stays outside it and fires. The mission plan for a Nishith strike against a defended radar installation or command bunker does not include acceptable losses. It includes the cost of the missiles. Those are different categories of cost."

He looked at Faruqi. "My evaluation recommendation is full qualification for service entry. The accuracy specifications are met and exceeded. The warhead options are appropriate for the target types the IAF will be assigned. The designation coordination requirement — the two-aircraft team with the SI-75 pod — needs dedicated training and dedicated doctrine, because the designating pilot and the strike pilot need to rehearse the coordination until it is automatic. That training programme needs to be written into the qualification requirements before the weapon enters service."

He sat down.

Bhatia presented the Sudarshan conclusions because Faruqi had asked him to, for the same reason Faruqi always gave the hardest analytical task to the person he believed could carry it most clearly.

"The Sudarshan passes all three qualification tests," Bhatia said. "I want to note the third shot's margin of 0.8 metres without dwelling on it beyond what it deserves. A pass is a pass. The weapon is qualified. What the 0.8-metre margin tells us is that the C-band frequency-shift adaptation is a technical area where further development is warranted, and the evaluation report should include that note clearly so the weapons development team knows where to focus the next iteration."

He paused. "The strategic significance of the Sudarshan entering service is worth stating plainly for the record, because the implications extend beyond the weapon itself.

"Every air defence system that threatens IAF aircraft depends on radar. The radar is the thing that sees the aircraft and guides the missile. Remove the radar and the missile is blind ordnance. The Sudarshan, in the hands of an IAF pilot, creates the following situation for any enemy radar operator: if he turns his radar on, he is helping the Sudarshan find him. If he turns his radar off, his own SAM system cannot engage the aircraft attacking him. He has two options. Both options degrade his effectiveness. There is no option that lets him both survive and do his job.

"The practical consequence of that dilemma, over the course of an air campaign, is that enemy radar operators begin operating their systems less — shorter transmission bursts, faster shutdowns, irregular patterns — all of which reduce the radar's effectiveness at its primary job of air defence. The Sudarshan does not even have to be fired to impose a cost. The knowledge that IAF aircraft carry the Sudarshan degrades enemy radar discipline in ways that benefit every subsequent strike mission, whether or not Sudarshan is employed."

He looked at the room.

"Together, Shefali, Nishith, and Sudarshan describe a different IAF from the one that flew in 1965 and 1971. In 1965, an IAF pilot attacking a defended target had to enter the threat envelope to hit the target, had to get behind the enemy to fire his missile, and had no weapon specifically designed to suppress the defences before the attack. In 1976, with these three weapons in service, none of those statements is true. The Sudarshan kills the radar before the strike begins. The Nishith kills the target from outside the defended zone. The Shefali kills any interceptor that survives and tries to intercept from any direction." He paused. "The three weapons work together. The doctrine that exploits them must be a combined doctrine from the start, not three separate systems bolted onto existing thinking."

Faruqi looked at him for a long moment.

"Agreed," he said. "The combined doctrine development starts next month. Three weapons, one framework, all adversary scenarios. Qualification reports on my desk by Friday."

He looked around the room.

"One last question, and I want each of you to answer it. What does the IAF look like, flying these weapons, that it did not look like before today?"

A silence.

Randhawa spoke first, because he had been sitting with the question since the Nishith's penetrator round detonated inside the concrete bunker. "It looks like an air force that doesn't accept that hitting the target means dying on the way in. It looks like an air force that separated those two things."

Walia said: "It looks like an air force where the strike pilot's job is to get the missile to launch parameters and come home. And then the missile does the part we used to do at the cost of our lives."

Nair said: "It looks like an air force where getting into the close fight with an enemy pilot is not necessarily dangerous, even if he has good missiles. Because whatever angle he thinks is safe — the crossing pass, the defensive turn — might not be safe anymore. The pilot who carries Shefali has made the other pilot's safe space smaller."

Bhatia said nothing immediately. Then: "It looks like an air force that was built for the fight we're going to be in, rather than the fight we were in the last time."

Faruqi closed his notebook.

"Debrief closed. Reports by Friday. Combined doctrine series commences fifteenth of October."

After the team had dispersed and the range was going quiet with the approach of evening, Dr. Venkat sat alone in the briefing room with her technical notes. She was writing the responses to the three qualification caveat items — Shefali's high-G tracking lag, Nishith's penetrator angle requirement, Sudarshan's C-band adaptation delay — because the technical responses were her work and she preferred to begin them while the day's data was fresh.

She was forty-one years old and had been an engineer for seventeen years, two of those with ISMC before joining Shergill Aviation's weapons division in 1972. She had spent the two years at ISMC on semiconductor process development, which had been valuable work, precise and demanding and important in its own way. But the weapons work had taken the precision of the semiconductor work and given it a different kind of weight. The things she built flew. The things she built decided outcomes in ways that process yield curves and deposition uniformity data did not. She had understood this intellectually from the start. She understood it differently after a day like this one — after watching Walia stand up and describe what a pilot accepting planned losses looked like, and what a pilot who no longer had to accept planned losses looked like, and the distance between those two positions.

She wrote the Sudarshan C-band response first. The hardware modification for the next development iteration was already designed — a receiver change that would reduce the C-band adaptation time from 0.6 seconds to 0.2 seconds, cutting the positional error on that shift from 3.1 metres to approximately one metre. The modification was in the pipeline for Sudarshan Mk.2. She documented it, referenced the test data, noted the timeline.

She wrote the Shefali high-G tracking response next. The solution was a software parameter adjustment to the tracking algorithm, not a hardware change. Her team had identified it three weeks ago. She documented it, noted it could be applied to current production units through a field update, referenced the test data.

She wrote the Nishith penetrator angle response last. The angle requirement was met by the current guidance software. The note to the evaluation report was simply: specification met at 63 degrees, within the 50-70 degree window. No action required. She wrote it and closed the folder.

Outside, through the narrow window in the briefing room's east wall, the Thar Desert was going through what it did every evening in late September — the specific transition from the last of the daylight into the blue-black clarity that the desert produced after dark, when there was no city within two hundred kilometres to put light into the sky and the stars appeared with a density and sharpness that cities made impossible. She could see a strip of it through the window.

She thought about the three weapons and about the men who had flown them today and about what the weapons gave those men that they had not had before. She thought about the S-27 Pinakas lined up on the apron and about the S-35 Tejas-Ms parked beside them, the aircraft that had carried the Nishith and the Sudarshan today, the aircraft that had been built in India by Indian engineers and Indian workers in Indian facilities, firing weapons that had been designed and built in India by Indian engineers, evaluated by Indian pilots who had been thinking all day about what Indian air combat doctrine should look like.

She thought about what it was to build things that worked. Not theoretically worked, not worked on the test bench, not worked in the simulation — actually worked, in the air, on the range, under the stress of real pilots in real manoeuvres in real conditions.

She picked up her bag.

She turned off the light.

On the south range, where the concrete bunker had stood that morning, there was a crater.

The desert was quiet.

The stars were very clear.

The work continued.