A Sensor Classification Scheme

We discuss a flexible and comprehensive categorizing scheme that is useful for describing and comparing sensors. I N virtually every field of application we find sensors that transform real-world data into (usually) electrical form. Today many groups around the world are investigating advanced sensors capable of responding to a wide variety of measurands. In an attempt to facilitate comparing sensors and obtaining a comprehensive overview of them, we present here a scheme for categorizing sensors. Sensor classification schemes range from the very simple to the complex. Extremes are the often-seen division into just three categories (physical, chemical, and biological) and the finely subdivided hierarchical categories used by abstracting journals. The scheme to be described here is flexible, intermediate in complexity, and suitable for use by individuals working with computer-based storage and retrieval systems. It is derived from a Hitachi Research Laboratory communication. Tables I-VI, containing possible sensor characteristics, appear in order of degree of importance for the typical user. If we take for illustration a surface acoustic-wave oscillator accelerometer, these entries might be as follows: the measurand-acceleration; technological aspects-sensitivity in frequency shift per g of acceleration, shortand long-term stability in hertz per unit time, etc.; detection means-mechanical; sensor conversion phenomena-elastoelectric; sensor materials-the key material is likely an inorganic insulator; and jields of application-many, including automotive and other means of transportation; marine, military, and space; and scientific measurement. Table I lists alphabetically most measurands for which sensors may be needed under the headings: acoustic, biological, chemical, electric, magnetic, mechanical, optical, radiation (particle), and thermal. A convention adopted to limit the number of Table I entries is that any entry may represent not only the measurand itself but also its temporal or spatial distribution. Thus, the entry “Amplitude” under the heading “Optical” could apply to a device that measures the intensity of steady infrared radiation at a point, a fast photodiode detecting time-varying optical flux, or a camera for visible light imaging. Manuscript received August 29, 1986; revised October 27, 1986. The author is with the Department of Electrical Engineering and COmputer Sciences and the Electronics Research Laboratory, University Of Galifornia, Berkeley, CA 94720. IEEE Log Number 8612468. With a particular measurand, one is primarily interested in sensor characteristics such as sensitivity, selectivity, and speed of response. These are termed “technological aspects” and listed in Table 11. Table I11 lists the detection means used in the sensor. Tables IV and V are of interest primarily to technologists involved in sensor design and fabrication. Entries in Table IV are intended to indicate the primary phenomena used to convert the measurand into a form suitable for producing the sensor output. The entries under “Physical” are derived from the interactions among physical variables diagrammed in Fig. 1. This is a modification and simplification of the diagrams used by Nye [ l ] and Mason [2] to show binary relations among the common physical variables. Most sensors contain a variety of materials (for example, almost all contain some metal). The entries in Table V should be understood to refer to the materials chiefly responsible for sensor operation. Finally, an alphabetical list of fields of application comprises Table VI. USES FOR THE CLASSIFICATION SCHEME A useful scheme should facilitate comparing sensors, communicating with other workers about sensors, and keeping track of sensor progress and availability. Categorizing might help one think about new sensing principles that could be explored, and Table I1 might serve as a checklist to consult when considering commercial sensors. All the entries in the tables have been given unique alphanumeric identifiers to facilitate use with computerized file systems such as electronic spreadsheets and databases used for storing information about sensors. The identifiers can be used as well in the keyword field of the lesserknown bibliographic utility refer, a part of the Unix operating system package, that enables a user to create and easily retrieve entries from a personalized database of citations to journal articles, books, and reports.