Airbags deploy when sensors detect a sudden, severe deceleration consistent with a crash, typically equivalent to hitting a fixed barrier at 8 to 14 mph or higher. The entire process, from the moment of impact to full inflation, takes just 20 to 30 milliseconds. That’s faster than a blink. But behind that split-second response is a chain of events involving impact sensors, a central computer, and a chemical reaction that converts solid material into gas almost instantly.
How the Sensors Detect a Crash
Modern vehicles use tiny electronic accelerometers, most commonly built with a technology called MEMS (micro-electro-mechanical systems), to measure how quickly a vehicle is decelerating. These sensors are small enough to sit on a circuit board but sensitive enough to distinguish a fender bender from a serious collision. When your car hits something, the accelerometer registers the rapid change in speed and sends that data to the airbag control unit, a dedicated computer usually mounted in the center of the vehicle.
Frontal airbags rely on accelerometers placed in the passenger compartment. Side airbags need a different approach. Because a side impact crushes the door inward with very little space or time to spare, sensors can’t sit far from the point of impact. Side-impact systems use either accelerometers mounted in the door area or pressure sensors that measure how quickly the door’s internal volume changes during a collision. Pressure-based systems have an advantage: they capture both how far the door deforms and how fast it’s deforming, which helps distinguish a crash from someone leaning against the door or closing it hard.
How the Computer Decides to Fire
The accelerometer alone doesn’t deploy the airbag. The airbag control unit runs an algorithm that analyzes the sensor data in real time, checking two key parameters: the change in vehicle velocity and the shape of the deceleration signal over time. The first parameter catches straightforward frontal crashes. If the velocity change exceeds a set threshold within about 35 milliseconds, the system fires the airbag regardless of crash type. If that threshold isn’t met right away, the algorithm keeps analyzing the signal pattern to determine whether the crash might be a pole strike or an angled impact, both of which produce different deceleration signatures than a head-on collision.
This distinction matters because different crash types look very different to a sensor. Hitting a narrow pole concentrates force on a small area, producing a sharp but initially slower deceleration compared to slamming into a wide barrier. Early electronic systems struggled with this. They either triggered too late for pole and angle crashes or triggered unnecessarily during low-speed bumps. Modern algorithms solve this by evaluating multiple parameters simultaneously rather than relying on a single deceleration reading.
The system also factors in whether someone is actually sitting in the passenger seat. A weight sensor integrated into the seat detects occupant presence. If the seat is empty or the occupant is too light (such as a very small child in a car bed), the system suppresses the passenger airbag entirely. The logic is straightforward: deploying an airbag into an empty seat wastes it, and deploying one toward a small child can cause more harm than good.
The Speed and Force Thresholds
According to NHTSA, frontal airbags are generally designed to deploy in moderate to severe frontal or near-frontal crashes equivalent to hitting a solid, fixed barrier at 8 to 14 mph or higher. That number can be misleading, though. A fixed barrier absorbs no energy, so hitting one at 14 mph is roughly the same as rear-ending a parked car of similar size at 28 mph. In real-world driving, many collisions that feel dramatic, like bumping a car at 10 mph in a parking lot, fall below the deployment threshold.
The threshold exists for a good reason. Airbags are single-use devices that cost at least $500 to replace, and they inflate with enough force to cause injury if they deploy unnecessarily. The system is deliberately calibrated to avoid firing during rough roads, minor bumps, or impacts with small objects like sign poles. Deployment is reserved for crashes where the occupant would genuinely benefit from the cushioning effect.
What Happens Inside the Inflator
Once the control unit decides to fire, it sends an electrical signal to an igniter inside the airbag module. This igniter sets off a chemical reaction that converts a solid propellant into gas. First-generation systems used sodium azide as the propellant. When ignited, sodium azide rapidly decomposes into nitrogen gas and sodium metal. The nitrogen gas is what fills the bag. The sodium metal, which would be dangerously reactive on its own, gets neutralized by additional chemicals (potassium nitrate and silicon dioxide) packed into the inflator. These react with the sodium to form stable, glass-like compounds called silicates, preventing the sodium from contacting moisture and producing corrosive byproducts.
The airbag fabric expands at 150 to 200 mph, reaching full inflation in roughly 25 to 40 milliseconds. By the time your body begins moving forward from the crash forces, the bag is already fully inflated and starting to deflate through vents in the fabric. That deflation is intentional. The bag absorbs your forward energy by letting gas escape in a controlled way, cushioning you rather than bouncing you backward.
How Seatbelts and Airbags Work Together
Airbags aren’t designed to work alone. Modern vehicles pair them with seatbelt pretensioners, devices that yank the belt tight against your body in the first moments of a crash. Pretensioners fire at a similar severity threshold as airbags and typically activate slightly earlier in the crash sequence, pulling you firmly into the seat before the airbag reaches you. This positioning is critical. An airbag inflating at 200 mph into someone who has lurched forward out of position can cause facial or chest injuries. The pretensioner’s job is to keep you where the airbag expects you to be.
Some newer vehicles are moving toward systems that detect your sitting position and adjust both pretensioner force and airbag deployment timing accordingly. If you’re leaning forward or sitting unusually close to the steering wheel, the system could theoretically modify how aggressively it deploys. This kind of adaptive restraint system represents where the technology is headed: not just detecting crashes, but tailoring the response to the specific person in the seat.
Why Airbags Sometimes Don’t Deploy
If you’ve been in a crash and your airbags didn’t go off, it doesn’t necessarily mean something malfunctioned. Several crash types fall outside the design parameters for deployment. Rear-end collisions typically don’t trigger frontal airbags because the occupant is pushed into the seat rather than thrown forward. Rollover crashes may not produce the specific deceleration pattern that frontal sensors are looking for, though curtain airbags with rollover sensors are increasingly common. Underride crashes, where your car slides beneath a higher vehicle like a truck, can bypass the front crush zone entirely, meaning the sensors never register the expected deceleration.
Low-speed crashes are the most common non-deployment scenario. A collision that feels jarring to you may still fall below the 8 to 14 mph barrier-equivalent threshold. In those cases, the seatbelt is designed to be your primary restraint, and the control unit correctly determines that deploying the airbag would do more harm than good.