A car crash that’s severe enough to trigger an airbag happens in roughly 200 milliseconds from first impact to the vehicle coming to rest — faster than a human eye blink, faster than any driver could consciously react. The airbag has to detect the crash, decide it’s real, and fully inflate inside a fraction of that window, which is exactly why the actual decision-making is handled entirely by dedicated sensors and a purpose-built logic system, with zero human input anywhere in the loop.
Two sensors, not one — on purpose
Modern airbag systems don’t rely on a single trigger, because a single sensor firing on its own creates real risk of a false deployment from a pothole or a minor fender-bender. Instead, the system requires two independent conditions to agree. A crush sensor, typically mounted toward the front of the vehicle, detects the physical deformation of a real impact. Separately, an arming sensor — an electronic accelerometer — measures the vehicle’s deceleration rate, and only registers as “armed” when that deceleration crosses a level that can genuinely only happen during an actual collision, commonly cited around 1 G or higher. Both conditions have to be satisfied within a tight window of each other — roughly 100 to 500 milliseconds — before the system will fire at all. That two-sensor requirement is the actual mechanism behind why airbags don’t go off during ordinary hard braking or a rough speed bump: neither one alone is enough.
The 5-inch, 30-millisecond rule
Once both conditions are met, the timing requirement gets genuinely extreme. Industry engineering standards work around what’s sometimes called the “5 inch, 30 millisecond” rule: the airbag must be fully deployed before an unrestrained front-seat occupant’s body — driven forward by their own momentum during the crash — travels five inches toward the steering wheel or dashboard. In practice, that means the full detect-decide-inflate sequence has to complete in roughly 10 to 30 milliseconds from the moment the arming threshold is crossed, with the physical inflation itself often happening in as little as 10 milliseconds once triggered. There’s no time in that window for anything resembling deliberation — the entire decision is a hard-coded threshold check running in dedicated hardware, not software making a judgment call.
Why this explains real-world deployment patterns
This two-sensor, threshold-based design is also why airbags sometimes don’t deploy in crashes that look severe from the outside, and occasionally do deploy in impacts that look minor — the system isn’t measuring visible damage, it’s measuring a specific deceleration signature the crush and arming sensors are calibrated to recognize as a genuine frontal collision. A low-speed impact against a soft object, or a crash angle the sensors weren’t specifically tuned for, can produce visible damage without ever crossing the deceleration threshold that arms the system.
The actual takeaway
Airbag deployment isn’t a single dramatic sensor “detecting a crash” — it’s two independent sensors that both have to agree, inside a window measured in milliseconds, before a hard-coded threshold check fires a decision with no time for anything resembling deliberation. The 5-inch, 30-millisecond engineering standard is the actual physical constraint the entire system is built around, and it’s a genuinely tighter deadline than almost any other safety system in a modern car has to meet.
