Airbag sensors

Citation

Monkman, G. (2005), "Airbag sensors", Sensor Review, Vol. 25 No. 2. https://doi.org/10.1108/sr.2005.08725baa.002

Airbag sensors

Airbag sensors

Keywords: Airbags, Sensors, Automotive

Few sensors have made such an impact on the automobile industry as impact sensors for airbags. The story started over two decades ago with experiments using simple accelerometers and tilt sensors. These were initially implemented only for anti- theft and intrusion alarms. Unfortunately, as anyone who lives or works near a car park knows, such simple devices have an alarmingly high “false alarm” rate. Such unreliability would be likely to cause, rather than limit, accident damage were such detectors to be used in airbag technology. Of course, a simple single dimensional measurement of g-force is not reliable enough to be decisive in triggering an airbag in a real accident whilst at the same time being capable of ignoring sudden accelerations and retardations which do not constitute a crash. Furthermore, the sensor output signal must be produced, transmitted and electronically processed within about 40ms. This was a knock out criterion for many of the (then) conventional accelerometers.

Although in the early days almost every known principle of accelerometry was tried, most of the deployment timing problems were only solved with the advent of microsensors which no longer relied on large masses and electromagnetic transduction. Despite the availability of these much faster microsensors, it was quickly discovered that only where a plurality of such sensors (known as “satellites”) were implemented, connected to a common control unit, was it possible to truly evaluate the seriousness of an impact. This evaluation is the all important criterion for the air bags go/no go decision. Furthermore, the positioning of the sensors was found to be critical and usually specific to the vehicle design. Each model of vehicle must be crash tested in order to determine the “crash profile” and with the help of computer simulation and modelling the optimal sensor location could be determined. Further developments then led to the addition of more sensors, specifically for side impact detection, which must also be integrated with the central electronics.

With such a large dispersion of sensors, the automobile manufacturers were then confronted with yet another problem. Owing to the dynamics of an impact, any modification to the vehicles structure or framework can result in changing the calibration parameters, thus compromising the reliability of the airbag system. This was of course, unpopular with manufacturers who wished to introduce a mutiplicity of structural variations within a given model range. However, market forces (or ever tightening safely regulations) prevailed and within the past decade the airbag has become a standard fitting to all new cars.

The rapid rate of development of air bag systems has placed enormous demands on the integrity of new sensors, all of which must be rigorously tested – product liability leaves little margin for error. Testing is traditionally carried out using vibration systems known as “shakers”. For vibrations up to 100Hz over several centimetres there are many electromagnetic and hydraulic systems on the market. Unfortunately these are rather limited as crash tests reveal sensor spectral outputs often containing frequencies of several kHz. Actuators capable of delivering strokes in the region of half a meter at frequencies of up to 1,000Hz are not inexpensive.

The latest problem facing the automobile industry is the need to differentiate between the impact from another vehicle – a situation in which both drivers are in danger and the airbags must be released, and that of an impact with pedestrians or cyclists in which case the drivers airbags should remain unused. However, additional devices must be deployed in order to protect the victim who usually ends up somewhere on the bonnet, or in worst cases, summersaults over the vehicle completely. In the same way as reactive armour on tanks can be used to explode and repel the effects of an incoming shell, so can airbags under the bonnet be made to react and cushion the impact with a human. More sophisticated systems employ additional airbags under the windscreen, or along the windshield pillars, intended to cushion secondary impact resulting from a pedestrian rolling over the bonnet. Given that over 20 per cent of all road accidents involve pedestrians, the new EEC regulations being brought into force in mid-2005 are perhaps not before time.

Gareth MonkmanBased at Fachhochschule Regensburg, Regensburg, Germany