In this paper, phase-modulated fiber-optic sensors are considered for use in acoustic, seismic, magnetic, pressure and temperature measurements. Specific examples, some test results, and advantages and disadvantages for each are discussed.

For more than 10 years fiber optic sensors have been projected as a dominant sensor technology for the future. This projection has yet to be realized. There continues to be a substantial level of research and development in various fiber sensors concepts yet only a few products are available commercially. The six major factors limiting the success of the fiber sensors in the marketplace are identified. An alternate approach to fiber sensors is described which take a systems approach to producing fiber sensors. It combines the advantages of optical fibers with the low power electronic sensor technology currently available. A single multimode fiber is connected between the readout and the transducer. Optical energy transmitted from the readout is converted to electricity at the transducer. Digital data from the transducer is synchronously transmitted over the fiber to the readout. Prototypes of magnetic (i.e., current), temperature, position, and proximity sensors have been built and will be described. A discussion comparing optical and electrical sensors is also included.

Modern photoelectric proximity sensors typically employ the use of a pulse modulated IR beam, a photodetector, amplifier, synchronous demodulator, and comparator to generate an output signal that indicates the presence of an object in the sense field (Figure 1). While these types of photoelectric controls generally work to acceptable performance levels at reasonable cost, they generally require a manual adjustment and high gain optics to achieve reasonable sensitivity and dynamic range. Adjustment is usually performed by setting amplifier gain; sometimes this is performed with an AGC circuit. The use of an amplifier gain adjustment has the drawback of limiting sensitivity and dynamic range under many normal operating conditions.

Modern industries around the world are daily facing problems associated with increasing their speed of production of piece parts and intermediates while maintaining quality and cost goals. Rugged, high speed sensors are readily available for sensing physical quantities such as temperature, pressure, alignment and the like. Less traditional parameters such as color, taste and texture have been left to off-line, human inspection. In the case of color, this has usually been a result of the speed, cost, operator complexity and fragility of existing equipment. This paper will attempt to review some current color sensing technologies and present some results obtained with a dispersive, high-speed Color Signature Sensor currently being developed at Honeywell's Sensors and Signal Processing Laboratory.

An inexpensive non-contact sensor for measuring the velocity of a broad class of targets, including liquid suspensions and diffusely scattering surfaces, has been developed and field tested. The operating principle is based on (1) illuminating the point of observation with two coherent beams, incident at different angles, (2) detecting light from both beams which has been scattered in a common direction, and (3) observing the interference between these two scattered beams as a function of time. When the target moves, the phase of the interference signal changes at a rate which is proportional to its velocity, thus allowing the direct measurement of the instantaneous surface or flow velocity. In the past, Laser Doppler Velocimetry (LDV) techniques have been largely limited to situations where single small particles pass through the illuminated region, generally giving rise to clean signals with high modulation efficiency. In this paper we describe an extension of this technique to the measurement of aggregates of particles in liquid suspension, or even diffusely scattering solid surfaces. In these cases, the interference signals are weak and are superposed on large background levels which fluctuate randomly with time. Accurate detection of these signals requires appropriate optical geometries and data processing techniques. We describe the design of such a sensor, and present test results. A measurement stability (precision) of better than 0.1% has been achieved in tests with paper pulp, and the indicated velocity correlates very well with the data from a bulk flow sensor. We use a laser diode source, PIN-diode detectors, inexpensive optical components, and an efficient digital data processing system: this leads to a sensor which is both compact and low in cost.

The use of reliable and accurate hygrometers to monitor and control the operation of drying hoods would result in significant energy savings and improvement in product quality. We have developed such a hygrometer, VaporSense, which operates at high temperature and humidity in environments containing high concentrations of particulates and corrosive vapors. VaporSense uses the differential absorption of light by water vapor wavelengths to determine absolute humidity. A UV fiber-optic prulit, is inserted into the contaminating environment. Optical surfaces are kept clean by a curtain of dry air. VaporSense prototypes have been installed in a variety of industrial drying sites. Stable, long term, contamination free operation was demonstrated.

Laser excited fluorescence of pulping liquors was investigated for use in the pulp and paper industry for process measurement and control applications. Liquors from both mill and laboratory cooks were studied. A Nd-YAG pumped dye laser was used to generate the excitation wavelength of 280 nm; measurements were also performed using a commercially available fluorometer. Measurements on mill pulping liquors gave strong signals and showed changes in the fluorescence intensity during the cook. Absorption spectra of diluted mill liquor samples showed large changes during the cook. Samples from well controlled and characterized laboratory cooks showed fluorescence to be linear with concentration over two decades with an upper limit of approximately 1000 ppm dissolved lignin. At the end of these cooks a possible chemical change was indicated by an increase in the observed fluorescence intensity. Results indicate that lignin concentrations in pulping liquors can be accurately determined with fluorescence in the linear optical region over a greater dynamic range than absorption spectroscopy. Laser induced fluorescence may also provide an indication of chemical changes occurring in the lignin structure during a cook.

Results of a study directed toward using observed spectroscopic features for the measurement of temperature in the combustion zone of recovery boilers are described. Emissions of the potassium doublets at 404 and 766 nanometers (nm) have been observed in recovery boilers and temperature and self absorption effects on lines shapes have been modeled. Predicted emission line shapes are strongly dependent upon predicted concentration values of potassium. Proper selection of concentration ranges results in good qualitative agreement of predicted line shapes with those observed in boilers and laboratory flame experiments. These results indicate that the temperature dependence of potassium emissions is complicated by self-absorption effects which limit the optical pathlength over which emissions are practically observable. Temperature measurement may be feasible using pattern recognition methods coupled with algorithms based on an emission model and realistic estimates of the emitting species concentration.

Substituting infrared thermal sensors for conventional thermocouples to measure the temperature of a product, or a point in a process, often provides the industrial user with distinct advantages such as freedom from contact with the product and better speed of response. The major disadvantage has always been higher sensor cost. Now that costs of ir sensors have come down, the non-contact approach is becoming more of a valid alternative, and the instrument or process control engineer often weighs the relative advantages of the two approaches before making a decision. With the advent of "smart" thermal scanning systems, however, it is becoming possible to rapidly measure and control several, many or all points on a product surface remotely and without contact, a capability without precedent, and not feasible with conventional contact sensors. This paper will trace the evolution of infrared noncontact temperature measurement, its development as a process control tool and the introduction of IR line scanners and imagers as industrial control sensors. Several applications of modern closed-loop control systems based on infrared sensors, scanners and imagers will be reviewed. 1. INTRODUCTION Temperature and thermal behavior of materials and fabricated parts in process are most critical factors in the manufacturing process. For this reason temperature is by far the most measured quantity in industrial process monitoring and control. Conventional methods of temperature measurement using thermometers and thermocouples are commonly used for the majority of monitoring and control applications. Non-contact temperature measurement using infrared sensors has become an increasingly desirable alternative over conventional methods as ir sensors have become less expensive, more reliable and electrically interchangeable with conventional thermistors and thermocouples. Now, with the introduction of innovative computer hardware and software, full image thermal control of products and processes is being made possible.

Optical sensor technology for arc welding control is critically reviewed. The basic sensing and control problems associated with arc welding are presented with joint tracking as the primary emphasis. The problem of imaging of the weld area and design of optics is considered including the unique coaxial optical system. General versus structured illumination is reviewed, with the various approaches described. Conclusions are drawn relative to the present state of the technology and expectations for the future.