Smart move for intelligent sensors

Citation

White, N. (2001), "Smart move for intelligent sensors", Sensor Review, Vol. 21 No. 1. https://doi.org/10.1108/sr.2001.08721aaa.002

Smart move for intelligent sensors

Smart move for intelligent sensors

Neil White

Keywords Smart sensors, Automotive

Almost four years ago, I was invited to be Guest Academic Editor for a special edition on intelligent sensors. Déjà vu or what? Any subject deemed worthy of its second special issue must be popular if not fashionable. But there again, modern-day life has Joe/Joanne Public using his/her Smart Mouse with his/her Smart Board PC or communicating with his/her Smart Talk mobile phone what a Smart A*** (of course Alec should be used here, but political correctness requires Alex).

The field of intelligent sensor technology continues to mature both in terms of the advances in the characteristics of the sensing elements themselves and also the electronic processing of data. Moore's law, for example, predicted that the number of transistors on a chip would double every 18 to 24 months from the late 1960s onwards. In 1995, chip population densities had reached ten million transistors per chip and the indications are that the 100 million figure will be passed within the next year. The rapidity of these developments means that the degree of sophistication within the processing block of the intelligent sensor also continues to advance at an ever-diminishing cost. Another important consideration is that of power consumption, and it is here that CMOS technology excels allowing the production of low-power intelligent sensor microsystems.

The area of microelectromechanical systems (MEMS) is a relatively new field for making miniature sensors and actuators using integrated circuit (IC) fabrication and related techniques. A majority of the research in this area has been undertaken within the last decade or so. As similar techniques are used for the production of both MEMS devices and integrated circuits, the predictions are that MEMS will eventually offer low-cost solutions to a wide variety of sensor (and actuator) problems. The limiting factor will be dictated by the overall complexity of the MEMS devices, as these require much greater processing time, access to very specialised equipment, complex packaging and sophisticated testing methods thereby leading to poorer yields than those obtained for IC devices. The design philosophy for future intelligent sensors will therefore require relatively simple sensing elements with dedicated electronic processing that compensates for the sensor deficiencies (offset, noise, non-linearity cross-sensitivities, manufacturing tolerances etc.).

Accelerometers are a case in point. Until recently the use of acceleration sensors on a car would have been prohibitive owing to the high costs associated with such devices. Modern air bags are widely used within the automotive industry and are a relatively low-cost item. The accelerometers used for crash detection are manufactured using MEMS technology and are produced in such high volumes that the cost has fallen dramatically. These sensors are, however, relatively "dumb" and are prone to causing the air bag to inflate whenever an abrupt acceleration is detected, whether or not a passenger is present. Future systems can be envisaged where the sensors will be capable of identifying not only the presence of a passenger, but also their weight and size and hence adjusting the force of inflation accordingly.

The boundaries between sensors and instruments, which once seemed so firm, are now quite blurred. Processes that were once confined to physically large electronic instruments are now available within the sensor housing. Thus, a sensor is now regarded as a system that inputs information and serves a host system. A complete intelligent sensor may therefore comprise: a primary sensing element, amplification, excitation control, active feedback control, analogue filtering, data conversion, local digital information processing and external information procession such as data fusion, neural networks or self-evaluation techniques.

The current issue of Sensor Review brings together some recent advances with the area of intelligent sensors. It is particularly encouraging to note the wide industrial take-up and diverse range of applications.