Click for next page ( 52


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 51
Opportunities in Sensor Technology INS RODUCTION In automated control systems, sensors represent one of the critical technologies that determine performance and the level of automation that is achievable. Although often overlooked, sensors in many cases represent the critical enabling technology. The computer processor, which usually appears to be the heart of an automated system, in reality cannot provide performance beyond that dictated by the sensor performance. Failures often are traceable to the malfunctioning of some inexpensive sensor component. As the push toward automation continues, the demands on the types, performance, and cost of sensors will grow. Optical sensors appear to offer performance and cost advantages that will enable many new applications to become possible. While the sensor market itself is a modest $3 to 5 billion a year market in the United States, it is the sensor performance and availability that enable many applications to occur. The total market for automated controls is roughly $50 to 75 billion per year. Relatively inexpensive sensors are in many cases embedded in expensive control systems, and it is the performance of the sensors that determines the marketability of the total system. Sensor technology can be highly leveraged and thus has significant economic impact. The difficulty in the sensor marketplace, however, lies in the fact that the marketplace is highly fragmented and diverse. Sensors that are useful in aircraft may have only marginal utility in the chemical industry. Market development and penetration of new products are hindered somewhat by an inhomogeneous marketplace. In 51

OCR for page 51
52 PHO TONICS spite of this, optical sensor technology is viewed as an enabling technology with the ability to create new areas in automated control. There are many different types of sensors in use today,~~3 and the application in many cases determines the requirements of sensor performance. In process control applications such as are found In chemical processing plants or in power generation plants, it is desired to be able to determine fluid flow rates, liquid levels, temperatures, pressure, rates of mung, status of various components such as valves and switches, and electrical currents and voltages; to inspect remotely various pieces of equipment; and to monitor personnel status. Sensors to perform these functions must be tied together with a robust telemetry system capable of providing the required bandwidth and, in some cases, able to survive such adverse processes as electromagnetic interference (EMI) and corrosion. Today conventional copper wire serves as the conduit in a telemetry system and connects numerous sensors, many of which are incom- patible with each other and therefore require special interfacing units. In moveable platforms, such as automobiles, ships, and aircraft, sensors are used for status monitoring as well as performance determination and alarming. Stringent environmental constraints are placed on the sensors in many of these applications. For jet engine monitoring, high temperatures as well as severe EMI and space constraints are often encountered. Pressure, temperature, and flow sensors, among others, are subjected to conditions that accelerate degradation. Sensor telemetry also may dramatically affect the performance and survivability of a sensor suite.4 For example, damage control systems--which are commonly made up of smoke detectors and sensors for liquid level, temperature, and rate of temperature rise--often fail in a fire not because of damaged sensors but because of the rapid loss of the telemetry system when it is exposed to severe heat conditions. Robotic devices, as they become more autonomous, must be able to sense the presence of objects, manipulate these objects, and place them in the proper location. As an example, small, sensitive pressure sensors located remotely in robotic hands are an important part of the control system for the arm. The ability to reliably sense pressure at a remote location is currently inadequate, and device improvements are desired. In automotive applications, small inexpensive sensors must function in the presence of high temperatures and EMI and in small spaces. Improvements in pollution and engine monitoring devices are highly desirable. In exploration for natural resources, highly sensitive sensing devices are often used. In oil exploration, for example, long, sensitive arrays of acoustic sensors are used to locate promising geological formations. Echoes or transmitted acoustic probe signals yield valuable information about the composition and structure of underlying strata. In well drilling, gyroscopes are used to determine the direction of the drilling while other sensors attempt to determine the local

OCR for page 51
OPPORTUNITIES IN SENSOR TECHNOLOGY 53 well-hole geology. In this application, high temperatures and corrosive gases that affect device reliability are often encountered. Most of the sensors described above refer to devices that directly probe physical phenomena. Another very important sensor class detects visible or infrared radiation and is used for surveillance or position monitoring. In robotic vision, solid-state cameras are used to obtain data on object location, size, and orientation. The data are processed with various algorithms and are used to control the movement of robotic devices. Solid-state cameras and processors are widely used for monitoring security perimeters, for damage determination, and for autonomous vehicle control. Focal plane arrays, which are microelec- tronic chips with two-dimensional arrays of photodetectors, also are finding use in passive optical radars such as IR search and track (IRST) equipment, for satellite imaging applications, and potentially for autonomous equipment control. PHOTONIC SENSORS Two of the principal classes of photonic sensors are fiber-optic sensors and focal plane (FPA) array imaging sensors. Both are currently in development and commercially available and promise significant advances in capabilities over current approaches. Foreign competition in both areas has proven to be substantial, and foreign manufacturers are in an excellent position to dominate In the area of FPA in particular, since the operative technology involves micro- electronic integrated-circuit approaches. If an edge exists for the United States ~ FPA and fiber sensor technology, it is due in large part to military investment in these areas. FIBER SENSORS After 8 years of development, optical fiber sensors are begs ng to emerge as competitive devices for performing sensing tasks such as those required for aircraft engine and flight controls, for shipboard machinery and damage controls, for medical probing, and for industrial process controls. Fiber sensors operate by having the perturbation to be measured (e.g., temperature, pressure, or displacement) modulate light propagating in a fiber. This can be ac- complished by placing specially designed coatings on the fiber or by using the fiber to conduct light to and from some material, placed in the fiber path, that reacts to the perturbation to be sensed and modulates the fiber-guided light. Fiber sensors have proved to be accurate and capable of operation in harsh environments contaminated with high EMI, explosives, or corrosive gases. Because of these attributes, fiber sensors offer unique opportunities to solve

OCR for page 51
54 PHO TONICS FIGURE 5.1 Examples of fiber-optic systems applications for marine use. existing sensor problems encountered in many applications. Additionally, the compatibility of optical sensors with optical telemetry makes possible the development of all-optical, multielement sensor systems capable of supporting large numbers of high-bandwidth sensor elements while at the same time eliminating the requirement for transmitting electrical power to the sensor site from the monitoring site.4 This combination of compatible fiber sensors and fiber telemetry represents an intrinsic advantage over conventional electrical technology. In Appendix E, the operation of fiber sensors is detailed, and the state of the art outlined. An example of systems applications is shown in Figure 5.1. Fiber-optic hydrophores, gyros, and magnetometers make up several of the high-performance sensor types. Performance parameters of these and other fiber sensors are shown in Table 5.1. Fiber hydrophores constructed with fiber interferometers have demonstrated detection sensitivities equal to or better than conventional piezoelectric devices. One of the main advantages of fiber-optic sensors is the spatial versatility of the sensing head. For an element with 30 m of fiber, the element can be a 30-m-long straight, extended element or a compact golf-ball-size omnidirectional element. The versatility also allows for a planar type of configuration as well as shaded, extended elements. A number of hydrophores for evaluation purposes have been successfully tested at sea. Single-element phones display state-of-the-art hydrophore performance. There has been considerable interest in both the United States and Europe in hydrophore arrays. In this application, multi- element sensors (driven by a single laser) with various types of multiplexing have

OCR for page 51
OPPOR7UNITIESIN SENSOR TECHNOLOGY 55 been considered to reduce the cost of the array. As of yet, only small demon- stration arrays with a few sensors have been built. Two types of fiber-optic magnetic sensors have been demonstrated, the Faraday effect sensors, which are useful for measurement of large magnetic fields ~ ~ 1 Oe), and interferometric sensors utilizing magnetostrictive materials for high-sensitivity devices (<1 mOe). In typical sensor designs, the sensor fiber is bonded to a magnetostrictive material. As the magnetic field changes, the strain in the material is transferred to a strain in the fiber core. Minimum detectable DC fields of 10~ Oe and AC fields of 10~ Oe range have been reported, which makes their performance equal to or better than existing, competing room-temperature non6~ber magnetic sensors. Fiber-optic gyros fabricated with fiber Sagnac coils have provided performance in the laboratory equaling the state of the art as achieved by the best ring-laser gyros.s Nearly all gyro applications have relatively stringent size requirements ranging from volumes of 1 ft3 in airplanes to volumes of a few cubic inches in missiles. This makes packaging considerations important in any engineered device. One of the leading approaches used today is the all-f~ber technique. Since the light never leaves the fibers, discrete optical component- alignment problems are avoided, minimizing gyro degradation due to vibration or thermal cycling. Recently, a ruggedized packaged fiber gyro for oil well- logging applications has been developed. This gyro is designed to operate to depths in the earth of 2000 ft over temperatures ranging from 0 to 125 C and to be able to withstand shocks to 100 g and vibrations of 40 g. This gyro can detect 0.05 rotations in azimuth and 0.2 in inclination and illustrates the state of the art in fiber gyro packaging. There are many reasons why fiber-optic gyros are currently attracting substantial interest. Fiber gyros are all solid-state and have no moving parts, implying reduced maintenance compared to present-day spinning mass gyros. Fiber gyros appear to have better potential sensitivity performance than the corresponding theoretical limits for ring-laser gyros, as is indicated in Table 5.1. They also do not have many of the problems that have plagued ring-laser development, such as optical lock-in, which required mechanical dithering, and precision block and high-quality mirror fabrication. Finally, they are con- structed from inexpensive components and have the potential to be inexpensive devices when compared with other technologies. Counterbalancing these advantages is the fact that the fiber gyro is still in an earlier stage of develop- ment than the technologies with which it is competing. Engineering issues of dynamic range and drift remain to be satisfactorily resolved. Recalling that the ring-laser gyro, which is now enjoying ~ successful introduction into commercial applications, required a development period of two decades, one expects that the much newer fiber-optic implementation will probably require another 5 to 7 years before it, too, successfully competes in a commercial marketplace.

OCR for page 51
56 C ~ ~ ~ o ~ ~ o o C; ~ ~ 'TWO ~ ~ ~ C_ _ ~ .~ _ F _ ~ ~ ' ~ ~ ~ i 5 U. Nat o U. C) Cal Ct d 1 o U. Cal A_ d Cal C) ~d" Ct 5- Em .5 C) ._ O ~4 ._ o Ct ._ Cal a> O ~ Ct O C) ~ it, V) ~ O O O C) Ct CtS O C) ~ ~ 4- ~ ._ ~ ~ ~ O ~ ~ ~ ~ ~ ~ o o ~ ~ ~ C ,~ C-g o o o ~ l ~o. o o l o o ~ o ct c) 4) ~o ~ ~ 2 ~ E, ~ ~ E ~9 E o ~ ~ o 2 u 2 U -

OCR for page 51
57 O ~ 5 5 . . . ~ Cal ; - _ Ct _ ~ +1 C) .- V ~ O o ~ o ~ * ~ ~ ~ ~ o ~ o o' ~ ~ U. ~ ~ 3 ~ ~ ~ ~ - . ~ ~ ~ i ~ o is, ~ o ~ o ~ o ~ +, C ~ ~ ~ ~ Rev ~ ~ be be ~ ~ ~ ~ o ~ O o to ~ ~ ~ ~ g of ~ o ~ o ~ ~ +1 +1 +1 1 ' ' ~ ~ O ~ ~ O a e - ~ By, ' e i ;- ~o ~o ~ 0 ~ s: ~ C> O o c' O ~ 5 a ~ ~ ~ ~ D _ by, C' o 5 Ct C) ._ Cal C' Ct .O

OCR for page 51
58 ao ~ o v _' - ce - - o . - s: .~ . us - ~: .O ct s c - ' a) s O O ' ~ :: - C ) ~ o 0 5 5 0 ~ o 5 D ~ ~ ~ ~ ~ ~ g A ~ >; 2 2 e ~ ~ ~ ~ ~ O O ~ = ~ ~ ~ ~ ~ O ~ ~ ~ ~ ~ ~ ~ ~ ~ j <= if_ 5 ~ ~ D C O o C ._ ~ ~ ~ ._ D ~ 0) ~ ~ ~ D 5 g ^ g I ~ o -_ o ~ g 5 ~ <' ~5 ~ cs = 5& ' ~ ~ 5 u: ~ ~ ~ ~ 1 _ -g S g 2 E E _ ~ a _ a I ~ E y s 5- $- ~ o 0 ~ `, 5 ~ 5 O ~ ~ ~ C ~

OCR for page 51
s9 3' >~ ~ ;:'~ ~ ~ ~ ~ ~ 3 o C) o ~ ~ ~ o O ~ ~ C A. o C O i,, ~ 'o ~ _ G o ~ .8 . o 4 D , _ i ~ a . , 8 0 ~ ~ , ~ 2,, ' ~ E E . sot on - 0 ~ cat I:: ~ CQ ~ C) ~q O c.) 0 0 ._ ~ ~ ~ ~ 0 > ~ _ ~ _

OCR for page 51
60 PHO TONICS Fiber-optic sensors appear to be ideally suited for machinery and process monitoring and control. In control applications, extreme sensitivities are not generally required, whereas a premium is placed on cost, simplicity, and reliability. The military is examining the use of f~ber-optic monitoring and control sensors and has actively pursued the development of several sensor types. Functions important for aircraft and ship controls are control of surface displacement, rotation, torque, and speed. Although these functions are also important in commercial processing and control, the high-EMI environment of military platforms that results from extensive radar and radio communica- tion operations provides an excellent incentive for the development of f~ber- optic position sensors. At present both military and commercial companies are developing fly-by-light flight control systems. Linear and rotary position as well as differential hydraulic pressure sensors have been developed. Position sensors operate as optical shutters, providing reflection, transmission, or absorption of light supplied by a source fiber in accordance with a pattern or mask inscribed in the moving member of the position sensor. These f~ber-optic devices replace conventional resistive bridges or magnetic induction position sensors. The first industrial application of these devices will undoubtedly be in hazardous/explosive environments where the all-dielectric nature of fiber sensors enhances safety. As indicated, improved fabrication, packaging, and production techniques are required to produce cost-competitive fiber-optic position sensors. Oil-pollution-mon~toring sensors are important for shipboard as well as industrial processing control. Fiber-optic-based sensors have demonstrated substantial improvements in accuracy over conventional devices and have demonstrated the ability to distinguish between oil and solid pollutants. Fiber sensors have been installed on over 800 vessels.2 Fiber-optic control sensors appear ideally suited for machinery and process monitoring. They generally possess adequate sensitivity for those applications. Power plant equipment and heavy electrical machinery appear to be prime candidates for early usage. The possibility of building distributed fiber sensors and embedding fiber sensors in material promises to satisfy numerous long- standing requirements. Fiber sensors embedded in composites, for example, should provide strain-probing capabilities of parts in motion. Additional oppor- tunities exist in aircraft and ship flight and machinery control systems. The military has actively pursued the development of several sensor types, including control and monitor sensors (e.g., for damage control and fuel status). These are currently being evaluated and are expected to find application. Lower-sensitivity fiber sensors have been incorporated into numerous control system demonstrations, and many control-type fiber sensors are commercially available. Included in this group are temperature, pressure, flow, torque, current, liquid-level, and several other types of fiber sensors. These are

OCR for page 51
OPPORTUNITIES IN SENSOR TECHNOLOGY 61 described in greater detail in Appendix E along with other fiber sensors used in medical applications. SOLID-STATE IMAGING DEVICES Since 1970, extensive research and development have been devoted to metal-oxide (MOS) semiconductor technology utilizing charge-coupled device (CCD) and charge-injection device (CID) concepts, leading to a revolution in data handling and processing of radiation-induced signals. Significant advances in multiplexing and amplifying signals from optical and infrared detector arrays have resulted. Devices resulting from these advances are finding application in strategic, tactical, and ecological reconnaissance and surveillance, both from the ground and from the air, as well as in consumer products such as miniature video cameras for various uses. Silicon technology was the early basic technology used in charge-coupled devices, and the first imaging CCDs were sensitive in the spectral region of 0.4 microns to 1.0 microns. Other windows at 3 to 5 microns and 8 to 12 microns have subsequently become accessible using indium antimonide (InSb) and mercury-cadmium telluride (HgCdTe) material systems. Imaging for regions below about 2 microns can be carried out with ambient or active illumination, whereas for longer wavelengths, the objects' own thermal radiation provides the signal to be detected. Research with InAsSb strained-layer superlattices, gallium arsenide/aluminum gallium arsenide (GaAs/AlGaAs), and other materials for use in IR devices also shows promise. Charge-coupled and/or charge-injection concepts have yielded, and promise to continue to yield, significant improvements in the performance of large-scale integrated detector arrays. Substantial increases in sensitivity with savings in weight, size, power dissipation, and cooling requirements have been realized with these new arrays. Increased scene information, coupled with evolving improved signal/clutter-processing technologies, offers capabilities for autono- mous detection and classification that were heretofore unattainable. Area arrays of 256 x 256 element complexity are becoming available in all spectral regions, with larger visible arrays having already been demonstrated. These will replace linear, mechanically scanned arrays in many applications, thus eliminating several moving parts, reducing size, and increasing reliability. In the visible and near infrared, large, buried-channel silicon CCD arrays have been used in the space telescope. Silicon imagers now can incorporate both storage and time-delay-integration (TDI) modes, which increases sensitivity significantly. Monolithic platinum silicide Schottky barrier IR-CCDs have demonstrated important improvement in scene uniformity, with good near- IR sensitivities. Many of these devices have already found their ways into both military and commercial cameras. Because of these detector improvements, video cameras have shrunken in size during the last decade by over an order

OCR for page 51
62 PHO TONICS of magnitude, a development that has opened up a vast consumer market to these products. In the infrared, scanned linear detector array technology continues to dominate. However, with continued progress in IR area staring arrays, it is expected that these new arrays will replace many of the scanning systems, thus eliminating the need for a conventional scanning mechanism and simplifying focusing optics. However, to compensate for array nonuniformities, correction on a pixel-by-pixel basis must be incorporated. Additionally, detector responsivity and readout nonlinearities require increased computations to correct for these flaws in an array output. These arrays, therefore, need to be closely coupled with a solid-state processor in order to perform properly. It is becoming clear that, as CCD/CID technology evolves, sophisticated, autonomously operated devices become possible, which in the future will affect many systems. For the first time, focal plane arrays permit image generation in a small, efficient sensor head. These imagers, when coupled with the appropriate processors and algorithms, permit target detection in clutter and object identification. These capabilities are becoming pacing requirements for advances in robotics, autonomous vehicles, surveillance, and data collection. Small, inexpensive imagers/processors would permit widespread use in applications where the size, orientation, and position of various objects are desired. Passive IR search and track radars are required for ship and aircraft navigation and protection, whereas small, smart seeker heads are required for guided munitions, as well as for a variety of consumer products, such as collision-avoidance devices and home protective systems. ENABLING TECHNIQUES Fiber-optic sensors have been in development for several years, and adequate performance has been demonstrated. One of the barriers to commercialization has been the diversity of the sensor market. While several fiber sensors are commercially available, compatibility among the sensor types has not been realized, and serious packaging of laboratory devices has occurred in only a few cases. Most of the fiber sensors listed in previous sections perform adequately in laboratory environments; however, drifts in calibration, due to environmental changes and other problems, have been encountered when poorly engineered or packaged devices have been fielded. It appears that all the components needed to fabricate fiber sensors are available, so that future development efforts can concentrate on making commercially viable sensors. Compatibility of packaging of several types of fiber sensor fill lead to the ability to multiplex multiple sensors on a common optical fiber. This in turn will lead to cost and performance advantages that will make fiber sensors the technology of choice. The strength of optical fiber sensors lies in the ability to use a

OCR for page 51
OPPORTUNITIES IN SENSOR TECHNOLOGY 63 common technology base to sense numerous physical parameters. Automated manufacture of fiber sensor arrays and associated telemetry promises significant cost advantages over current practices. This fact, coupled with resistance to EMI and corrosive environments and competitive performance, should ensure the widespread use of this technology. As for the many other technologies emerging from U.S. laboratories, the difficulty in commercializing the technology rapidly relative to foreign competition is a major concern in the optical fiber sensor area. Projects to standardize fiber sensors, so that compatibility issues are resolved, and to integrate fiber telemetry are desirable. The military ~11, in all probability, perfect the sensitive surveillance types of fiber sensors, and these will diffuse into the commercial marketplace with time. Focal plane array development, on the other hand, requires more research to address fabrication problems even though many commercial FPAs are currently available. The FPA technology is still evolving, and continued research is required to develop better materials-growth approaches, interface improvements, fabrication yield improvements, and signal processing to support image acquisition and processing. Nonuniformity of detector response from pixel to pixel requires considerable post-processing for IR-FPAs and reflects the difficulties in materials growth and fabrication of the array. Array yields remain low, which reflects the evolving nature of designs and of processing approaches. Considerably better control over device parameters has been realized with visible FPA structures versus IR-FPAs. This is traceable to the use of highly developed silicon technology in the visible region of the spectrum--whereas InSb and HgCdTe material systems and processing techniques for IR-FPAs are less developed. Funding and incentives are needed to make focal plane arrays cheaper and more reliable. Of equal importance is the development of custom processing hardware and software to support focal plane array data acquisition and processing. The lack of very efficient image processing algorithms limits the ability to perform real-time feature extraction, clutter suppression, and related data manipulation and determines the size of the processor needed to perform a function. In many cases, a small sensor head with a volume of a few cubic inches creates sufficient data to require a VAX-size or larger computer for processing. The long-term goal, which is supported by the potential of the technology, is to develop algorithms and processors efficient enough to permit the processor to also fit into a volume comparable to that occupied by the sensor. This will open the door to many autonomous vehicle, robotic, and military applications.

OCR for page 51
64 PHO TONICS REFERENCES 1. Giallorenzi, T. G., J. A. Bucaro, A. Dandridge, G. H. Sigel, J. H. Cole, S. C. Rashleigh, and R. G. Priest. 1982. Optical fiber sensor technology. IEEE J. Quant. Elec. QE-18~4~:626-665. Pitt, G. D., et al. 1985. Optical-fibre sensors. IRE Proc. 132~4~:214-248. Jackson, D. A., and J. D. C. Jones. 1986. Fibre optic sensors. Optica Acta 33~12~:1469-1503. 4. Dakin, J. P. 1987. Multiplexed and distributed optical fire sensor systems. J. Phys. Eng. 20:954-967. 5. Bergh, R. A., H. C. Lefevre, and H. J. Shawl 1984. An overview of 6~ber- optic gyroscopes. J. Lightwave Technol. LT-2~2~:91-107.