Most of an air defense engagement is the decision not to fire

In short: an interceptor is the visible part of air defense, but the expensive engineering sits in the radar and the battle management computer that decide which rockets are worth shooting at. In a heavy salvo, the software ignores most incoming rounds on purpose, because they are already falling on empty ground.

Watch any clip of a rocket barrage and the story writes itself: rockets go up, interceptors go up, the sky flashes. The part nobody films is a computer sorting dozens of tracks in a few seconds and rejecting most of them. That rejection is the product.

קראו גם: The best tank armor on the battlefield stops nothing on its own · Spike Is Not One Missile, and That Is Exactly the Point · Iron Dome gets the headlines, but David’s Sling stops the threats that would actually level a building

The radar sees a trajectory, not a rocket

A launch is detected within seconds, but detection alone is close to useless. The system needs a predicted impact point, and it has to build one from a partial track of an object that is still accelerating. Radars in this class, including the EL/M-2084 family, run search and fire control at the same time while tracking many objects at once.

The output is a map, not an alarm. Every track gets a projected landing area with an error margin, and that margin is what drives the next decision. Wide margin over a city means fire. Tight margin over farmland means hold.

Why a system fires at a minority of the salvo

Rockets built in workshops are inaccurate by design, so a large share of any barrage lands in open areas. Firing at those rounds would drain the magazine in minutes and leave nothing for the rounds that matter. Published accounts of Iron Dome operations describe exactly this logic: engage what threatens people and infrastructure, ignore the rest.

That is also the honest answer to the price complaint. An interceptor costs far more than the rocket it kills, and the per-round comparison only works if you assume every rocket must be intercepted. The software exists so that assumption never has to hold.

Note the third row. Ranking is where a defense either holds together or collapses, and it is the reason a battery is measured by how many simultaneous engagements it can manage rather than by missile range.

Layers exist because one sensor cannot do it all

A short range radar tuned for a rocket at ten kilometers is the wrong instrument for a ballistic threat entering from high altitude. Different threats need different sensors, different intercept geometry and different warheads, which is why one continuous dome over a country was never the architecture. Each layer hands the picture to the next.

The same constraint limits newer hardware. A high energy laser still needs the same track quality and the same classification decision before it points anywhere, and it needs clear air on top of that. Cheap shots do not remove the need for good data.

What actually limits a battery in a bad hour

• Number of engagements the fire control can hold at once, not total missiles in storage.
• Time to reload launchers between waves, measured in minutes under pressure.
• Communication links between radar, command post and launchers spread across kilometers.
• Trained operators who can override an automated ranking when it is clearly wrong.
• Debris risk over roads and buildings, which sometimes rules out a valid shot.

Operators also carry a duty that no software has: releasing a shot they know will scatter fragments over an inhabited street, because the alternative is worse. That call stays human.

If you want a useful way to read the next barrage report, skip the interception percentage. Ask how many rockets were classified as non threatening and how many launchers were available in that sector. Those two numbers explain the outcome; the highlight reel does not.