National Academies Press: OpenBook
« Previous: THE SYSTEMS APPROACH
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 34
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 35
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 36
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 37
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 38
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 39
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 40
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 41
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 42
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 43
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 44
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 45
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 46
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 47
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 48
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 49
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 50
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 51
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 52
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 53
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 54
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 55
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 56
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 57
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 58
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 59
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 60
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 61
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 62
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 63
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 64
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 65
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 66
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 67
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 68
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 69
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 70
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 71
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 72
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 73
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 74
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 75
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 76
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 77
Suggested Citation:"HUMAN FACTORS." National Academy of Engineering. 1973. Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation. Washington, DC: The National Academies Press. doi: 10.17226/18640.
×
Page 78

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

34 BIOMEDICAL AND SOCIOLOGICAL IMPLICATIONS OF U.S. TELECOMMUNICATIONS RESEARCH I. Introduction Research in telecommunications technology has attracted engineering talent of the highest calibre in this country and overseas. The record over the past quarter century has been one of unremitting pursuit of excellence. With such out- standing intellectual resources, and with such highly motivated individuals and teams, it is therefore to be expected that the telecommunication field itself would competently establish long and short-term goals appropriately related to national and international needs. That these internal judgments and assessments have been highly successful in promoting technological development is without question. What is now becoming less clear is the extent to which the application of current arts of telecom- munications has been most effectively carried out for general betterment over the widest range of economic strata within our own and other societies. In this context, the one question immediately takes on a dual and pressing character. On the one hand, there is the problem of diffusion into society of communications devices and systems poorly understood by the user. In turn, this lack of understanding breeds a paranoid reaction against acceptance of further technology, no matter how carefully devised or responsibly applied by expert innovators. On this problem, Murray Gell-Mann recently stated that "these are the unfortunate effects of care- lessly deployed or carelessly diffused technology." On the other hand, there is a separate but related goal of optimizing the quality of life for the majority of our citizens for as long as possible in their life span. This is a question clearly separate from user acceptance of a particular device or system. It is a question in which society's broadest and best evaluation of its own goals must encompass the technologies of communica- tions systems as such, and endeavor to build them into the most effective forms of needed human communications. Clearly, we have reached a position where all current forms of communication are not needed. Their very existence, or

35 at least their trivial application, has been and may continue to be detrimental and even disastrous to human society. Therefore, it is necessary in future planning that we discard or suppress those forms of communication which have adversely affected the quality, as well as the breadth and depth of human communica- tions. It is equally important that we nurture and foster communications techniques that promote our social development. We should seek by their implementation and growing sophistica- tion to prevent the tragedies in human relations that so commonly follow a failure of needed communications.' Health care delivery, biomedical research, and child and adult education are representative areas which will depend in- creasingly on the full gamut of telecommunications development to broaden and increment man's unaided capabilities. Many of these areas present problems that are unique to the United States. These problems are often complicated rather than helped by high levels of technology in the general community, and the associated "careless diffusion and careless deployment" cited above. Man's capacities in telecommunications reflect the special development of his hand and brain in toolmaking. In the past 20 years, he has strikingly augmented his capacities in per- ception and decision-making with the electronic computer. These new tools which extend man's senses are already inextricably interwoven with many forms of telecommunications; and as they have evolved, we have recognized major new areas of needed psychophysiological and telecommunications knowledge that would optimize man's interactions with computing devices. How well equipped is the U.S. to undertake such a sub- stantive range of tasks in the coming decade? In particular, granted our technical superiority in both telecommunications and computer sciences, does it necessarily follow that our society will apply them to the most effective solutions of socially pressing problems? Or will others make more effective use of U.S. technology? The following para- graphs will attempt an evaluation of some major issues, based on domestic and foreign developments.

36 II. Health Care Delivery This Administration has committed itself to the concept of universal health care. We are now clearly far removed from this desirable goal, in terms of medical manpower and of clinical facilities. Moreover, the training of more doctors may not of itself offer an economically feasible solution. Instead, we may anticipate development of teams of para- medical professionals, heavily dependent on modern telecom- munications, and entrusted with much of the health care delivery in both urban and rural areas. A. Logistics of Communication Systems in Health Care Delivery The unfortunate inhabitant of a city ghetto may be as isolated from medical care as those living in remote rural areas, if he were to rely on available forms of medical prac- tice. For both groups, prompt availability of medical care may be a matter of life or death. In city areas, traditionally served by large city or county hospitals, there will be trends toward many small clinics distributed through the community, and delivering "minimal health care" in areas where none now exists. These clinics would operate under the direct surveillance of quali- fied medical staff at a regional medical facility. Staffed by paramedical personnel, they will require two-way color video links with one or more central medical facilities, including a capability for transmission of light microscopy images from the field station. They will also have facilities for transmission of graphic medical data, including the electrocardiogram, respiration, electroencephalogram, and a variety of data on the muscular systems. In rural areas, there are numerous minority communities, in addition to isolated farming groups, for whom local medical care is non-existent, and a visit to the doctor entails a long and often hazardous journey by road. Here, too, future plans for medical care envisage both mobile vans and regional clinics operated by paramedical personnel in close communication with regional medical centers.

37 B. Principles of telemedicine The traditional examination by the physician has been heavily dependent on his perception with unaided senses of significant aberrations in structure and function. The great advances in biochemistry over the past 25 years have brought about a veritable revolution in ancillary tests that have added orders of magnitude to the precision of the clinical examination. Space age technology added a further dimension. Bio- medical data acquisition systems have made increasing use of microminiaturized transducers, signal conditioners, and bio- telemetry systems that are a direct spinoff from the space effort. Early experience with these techniques has indicated the feasibility of a remote medical examination. The arts of telemedicine will evolve slowly. Feasibility studies have involved a medical technician working under in- structions from a physician. We may anticipate bold steps into the use of telefactors, of types that will evolve from lunar and planetary exploration programs. These may ultimately sub- stitute in some circumstances for the medical technician in assisting the physician in his remote examinations. On the other hand, the major thrust of telemedicine programs will con- tinue to be by the extended use of paramedical professionals, a group for which needed training programs are only now emerging. C. Communications channels for telemedicine Telemedicine schemes will make full use of all current forms of telecommunications. Patient medical data, such as LKG, EEG, etc., can be effectively transmitted over voice grade telephone lines in an IRIG narrow band FM multiplex format. They can be demodulated in real time and subjected to further computer analysis and pattern-recognition techniques. Already, feasibility has been established for such trans- missions on a routine basis over international voice-grade telephone circuits. These same voice circuits are also in- valuable in their traditional role of patient or paramedic interrogation by the physician. Video information further enhances quality and quantity of medical information transfer. Where channels are restricted

38 to voice bandwidths, slow-scan techniques are thoroughly feasible, and sequential optical filtering with color photog- raphy now allows transmission of color images. Radio transmission in H.F. portions of the spectrum is already suited to these narrow band requirements for dis- tances beyond VHP or UHF coverage. The latter, however, will be increasingly used, both for direct communications between outlying fixed or mobile clinics and regional medical centers, and also in providing links to satellites. At UHF frequencies, high resolution two-way TV coverage can be planned. In summary, organizational schemes in telemedicine will remain flexibly related to varied capabilities of each geo- graphical unit — the intellectual capacity and training of lay or professional operators in the field, the channel capaci- ties in data transmission, and the sophistication of central or regional medical facilities and personnel. A substantial period of trials for prototypes of these systems will be necessary before blueprints can be offered that would assure high cost- effectiveness as well as efficient delivery of health care in typical urban and rural communities. We do not yet know the essential verbal communications that would be part of the examination protocol; nor do we know how best to use mechanical sensors as substitutes for the physician's eyes, ears, and hands. With increasing use of cable TV (CATV) systems, there have been proposals for use of these systems for transmission of medical data from the home to the doctor or hospital. Although there appear to be no insuperable technical obstacles, there may well be certain social problems. Transducing equip- ment that would transmit a major body of needed medical data would probably be expensive, require regular maintenance and calibration, and be used only intermittently. Moreover, the user would undoubtedly need training in its use, and these technical skills may well be beyond the capabilities of the elderly, the infirm, and those whose lives are typical of the majority in this country in being "technologically deprived" in terms of a reasonable understanding of instrumentation systems.

39 D. Specialized Telecommunication Needs within Hospital and Medical Center Environments Intensive patient care in medical and surgical emergen- cies, including the immediate postoperative phase following major surgery, has placed great emphasis on sophisticated laboratory analyses, as well as in monitoring vital signs. Much of the recognition of complex patterns in these analyses has become, and will be increasingly dependent on computer-based analyses. Video monitoring of walking patients, or as an adjunct to bed care, is increasing. For these reasons, a veritable net- work of internal communications is essential in hospital prac- tise. Shock hazards from multiple lead attachments, including low impedance paths through renal and vascular catheters, have led to use of radio biotelemetry and fiber optic techniques to eliminate dangerous leakage paths in intensive care procedures. The present state of the art may be regarded as intermediate between older "hardwire" transducing methods and future develop- ments in "no contact" sensing devices. As a genotype of possible future developments, devices of the type using Josephson junc- tions have shown great promise in remotely sensing magnetic components of cardiac contraction and brain activity. Not only will developments in this area avoid present problems of shock hazards, etc., but the improved mobility of the critically ill patient has significant therapeutic advantages and makes nursing care easier and more efficient. E. Comparative Evaluation of Overseas Developments of Telecommunications for Health Care Delivery There can be little doubt that, in the elements of sophis- ticated technology, the United States possesses both prototype techniques and fully evolved systems that are substantially in advance of similar developments elsewhere for health care delivery. On the other hand, there are important factors that may limit the deployment and efficient evolution of these systems in furtherance of health care in the U.S. Foremost is the extremely high cost of our existing health care system, in which expenditure has grown from $50 to $l00 billion in the past five years, and which may continue to grow exponentially unless it is

40 drastically restructured. Creative application of telecom- munications technology may hold reasonable hope of holding some of these costs in check; it may even contribute to reversing the current exponential trend. There is the further prospect that adaptations of this U.S. technology to medical care in overseas countries may pro- duce good models of system applications before they are avail- able in the U.S.A. For this reason, it may be desirable to view future developments on a collaborative basis with foreign investigators, rather than competitively. This may be a highly productive route to needed knowledge. It would be expected that some foreign applications would bear a unique stamp reflecting local needs, but that much of their new knowledge would be valuable by extrapolation to U.S. problems in medical care. Moreover, developments in satellite communication will clearly allow extensive medical data transmission over inter- national circuits, with the potential for use of sophisticated pattern recognition and image processing methods at U.S. medical centers. These techniques are not likely to be widely available abroad in the foreseeable future. III. Telecommunications Research in Child and Adult Education Classroom education has been relatively little changed by available telecommunication technology. It is an area lacking the unified approach now possible from new knowledge gathered in the separate fields of brain physiology, the psychophysiology of perception, and the arts of telecommunications and computa- tion. Studies in brain physiology have revealed substantive correlates, both for individuals and for groups, of the processes of alerting, attention, and in some degree of decision-making. Psychophysiological research has indicated some of the ways in which we may optimize the environment, and in turn, optimize the learning rate. Recognition of states of high attention, and of the waxing and waning of levels of alertness, demands interactive relations with sophisticated computing systems. For the instructor, these methods would offer an opportunity to regulate the informational flow and provide him with more positive links with class members than is now possible. From preliminary studies in these areas, exciting pros- pects are emerging. For the normal child, there is the

41 possibility that conventional teacher-student interactions will be usefully supplemented by complex audiovisual dis- plays that may be temporally segmented to meet cyclic changes in levels of subjective awareness. For mentally retarded children, numbering about l percent of the U.S. child popula- tion, there is the prospect that electrophysiological signa- tures of brief moments of attention may be well enough recognized to enable presentation of learning tasks at these times. To be effective, these methods applied to both normal and defective children will rely on complex two-way telecommunica- tions with the instructor. Our knowledge of the physiological signatures of brain actions continues to grow rapidly, but this growth is dependent on further major research in tele- communications and computing sciences. Much research in perception, models of brain action in the perceptual process, and research in interactive computer processing is now proceeding in the Soviet Union and in Soviet bloc countries, particularly in Czechoslovakia. On the biologi- cal side, this work is of high quality, but is technologically limited by available computational capabilities. It is backed by a long history of mathematical modeling of biological and psychological processes. IV. Man-Machine Relations that Extend Human Perception and Action The ability of the digital computer to function as a pattern recognizer has opened new vistas in man's social and environmental interactions. There remain substantial prob- lems, however, in the effectiveness of human communications with computing devices, and in man's ability to interpret complex computer output displays. In part, these relate to inherent differences in opera- tional modes in brain and computer; the brain functioning as a highly parallel processor, with relatively slow internal bit rates, and significant error probabilities that may relate to "noisy" or pseudorandom internal behavior; the computer characterized by a single address format and very high internal bit rates. These constraints in the computer may account for a quite rigid adherence to alpha-numeric displays that probably

42 offer a considerable informational mismatch in interactive processing with a human operator. Yet the alternatives in computer organization that might optimize human interaction are not immediately apparent. Nevertheless, the rewards for improvement of input-output organization would be sufficiently great that it would be well worthwhile to consider a direct "brain-computer interface," in which we might take advantage of a "perceptual shorthand" in the computed display format, as well as finding ways to relax rigorous descriptions in input data sets, as proposed in the theory of "fuzzy sets." Future developments may extend the capability of the com- puter as an adaptive filter in recognition of electric brain patterns. Preliminary studies have indicated its possible application in this way to control of a variety of prostheses. Clearly, future developments in telecommunications will envisage these very complex pattern recognitions as occurring at the terminals of long communications links, at first over terrestrial paths, and perhaps later over interplanetary distances in space exploration. The value of reliable com- munications links to these automated pattern recognizers and decision makers would clearly be paramount. W. R. Adey School of Medicine UCLA

43 EXCESSIVE COMMUNICATION The development of civilization has been marked by milestones in the art of communication. Such markers have included the development of language, its retention in written form through picturegrams and alphabets, the invention of paper which makes written information transportable and retainable, the art of printing which greatly multiplied knowledge available to all men, and more recently the contributions of telecommunications, the telegraph, the telephone, television, and the computer. Each of these developments has added to the quantity and the rapidity with which information is presented to the human recipients who are the ultimate users. A small proportion of information must be decoded and used by machines, but the amount presented to a machine is usually limited in the design of the system incorporating such machines, and hence is within their capacity to handle. Such is not the case with the present flow of information terminating on the individual human being. While telecommunications, i.e., the coding, processing, transportation and decoding by electrical means, represents only a portion of the total information processing in modern life, the developments of telecommunication theory and practice have made major contributions to the philosophy and action involved in all communication processes, including publications and just plain talk. Examples of telecommuni- cation theory which have carried over into other fields of thought have been the recognition of the universality of the concepts of feedback, signal-to-noise ratio, redundancy, coding and decoding and impedance matching to achieve opti- mization, channel loading, etc. The human intellect must be regarded as the most impor- tant receiver for most messages. Information theory has proposed that when any communication channel becomes over- loaded, communication deteriorates rapidly, or even ceases abruptly. Modern men, especially those charged with making the most important decisions, are being presented with information at an ever increasing rate. It is proposed that more effort should be applied to research to determine quantitatively just what are the limits in information processing inherent to human beings immersed in our social structure which floods them with oral, visual, written and printed information, some relevant (signals) and much irrelevant (noise). Information of this

44 type is particularly important when the recipient is expected to use this information to make appropriate decisions and take action, or when rapid decisions affect human safety and well being. Such research will require the correlation of several disciplines in the design and implementation of experiments which will involve psychological, social, and physical theory and practice. I have asked Dr. Stanley N. Roscoe, who is both a psychologist and an engineer to comment on the need for such research. His remarks entitled "The Determinants of Cognitive Channel Capacity," follow this article as Appendix A. His professional field involves particularly the problems of the instrument-pilot interface in the airplane cockpit. He has conducted some experiments which are described in his report, "Assessment of Pilotage Error in Airborne Area Navigation Procedures."* This is a "for instance" example of how such problems can be reduced to quantitative measurement. The details of his experi- ments excerpted from the report are attached as Appendix B. It is the theme of the report that many more experi- ments in the area of human communication overload should be developed and supported to determine the effects of overloading of the message channel in the presentation of information to human beings in the presence of noise, using the word "noise" in its broadest sense. A knowledge of the deterioration in the quality and quantity of usable infor- mation under these circumstances can have important social and economic values. W. L. Everitt Dean Emeritus College of Engineering University of Illinois, Urbana Attachments (2) *Prepared under the auspices of the Aviation Research Labora- tory, Institute of Aviation, University of Illinois-Willard Airport, Savoy, Illinois for the Air Force Systems Command (Technical Report ARL-72-24/AFOSR-72-l3, October l972) .

45 APPENDIX A THE DETERMINANTS OF COGNITIVE CHANNEL CAPACITY In the realm of intelligent behavior, each individual has a time-variant momentary cognitive attention capacity and, to varying degrees, the ability to distribute that capacity to attend to different matters as a function of time. In many respects this cognitive channel capacity can be likened to the channel capacity of a communication system; in some it can not. The differences between animate and inanimate channel capacity are most noticeable as the two types of systems approach saturation. When an inanimate communication channel becomes saturated, communication breaks down — the flow of information stops; when an animate communication channel becomes saturated, it exhibits more or less graceful degradation — some messages continue to get through. Typically human cognitive behavior approaches saturation in situations requiring the rapid time-sharing of attention among competing stimulus inputs having associated action pri- orities that vary as functions of time, ambient circumstances, or the changing objectives of the individual. Examples of complex activities requiring rapid time-sharing of attention to a wide variety of stimulus inputs are flying an airplane under instrument flight rules in a congested air traffic situation, calling defensive signals while linebacking for the Washington Redskins, or serving as President of the United States. In each of these examples, an abrupt breakdown in responses to stimuli could be tragic. The experimental study of animate cognitive channel capac- ity at or near its saturation level may lead not only to the psychological discovery of the determinants of human per- formance in complex stimulus-decision-response situations, but also may yield bionic transfer to the design of inanimate com- munication channels that exhibit graceful degradation. Stanley N. Roscoe Aviation Research Laboratory University of Illinois

46 APPENDIX B A COMMON METRIC FOR RNAV*EQUIPMENT CERTIFICATION The categorical assertion that a system's blunder prone- ness can be assessed by measuring a pilot's residual atten- tion while using it is based upon an experiment recently completed at the University of Illinois (Kraus, l972; Kraus and Roscoe, l972). The purpose of the experiment was to evaluate a reorganized manual flight control system in a realistic, complex mission environment. A Link GAT-2 gen- eral aviation trainer, interfaced with a high-speed digital computer, simulated procedural and performance characteris- tics of a representative airborne area navigation system operating in an IFR air traffic environment. In the experiment, three variables were manipulated, the type of manual control system flown, the waypoint storage capacity of the simulated RNAV computing system, and the level of side-task loading to which the pilot sub- jects were submitted. A saturating level of side-task loading served both to induce a stressful flight situation and to provide an inferential measure of each pilot's residual attention as a function of the equipment variables being investigated. The results of the experiment are reported in the papers cited; their relevance to the certification of equipment lies in the fact that each pilot's residual attention, as measured by the rate at which he could cope with the information- processing side task, varied in a sensitive, orderly, and statistically reliable manner with each change in equipment characteristics, despite the widely differing levels of residual attention exhibited by different pilots. The computer-controlled side task was automatically adaptive; the faster the pilot responded, the faster informa- tion inputs were presented. A horizontal array of transillu- minated numerals, 0 through 9, was mounted immediately above the primary flight group on the pilot's instrument panel. Numerals l through 8 were illuminated in a random sequence, and as each appeared the pilot could extinguish it by pressing the corresponding numeral on a scrambled keyboard mounted above his right knee and out of his normal field of view. During operation, when a pilot extinguishes an illuminated numeral, another appears after a 0.75-second delay; thus, if "Area navigation is any system of navigation that allows use of all the airspace, without restriction associated with the geographic locations of radio navigation facilities.

47 he were to respond to every stimulus with no delay, he would be processing a theoretical maximum of four bits of informa- tion per second (three bits per response). However, if a pilot fails to respond within six seconds, the stall warning horn starts sounding with rapid intermittence, calling his attention to the side task. The effect of the blasting horn can be dramatic. If difficulty with primary flight control or confusion over RNAV procedures were the cause of the pilot's original failure to respond within six seconds, the onset of horn blasts can be followed successively by annoy- ance, frustration, hostility, and panic. In the experiment, the stress created by the side task was accompanied by a doubling of the frequency of pilot blunders, regardless of the manual control system in use. Despite the stressful effect of the elevated task loading and the four to one range of residual attention among pro- fessional pilots (approximately 0.25 to l.00 bits per second), well-designed systems approached freedom from blunder prone- ness, indicating that it would not be unreasonable to require demonstration of a specified level of blunder-free residual attention by a group of properly qualified pilots for RNAV system certification. The particular information-processing side task described is only one of many that might be employed. It was used because it was simple to implement and score and because it was found to work during preliminary experimentation. A more complex cross-adaptive logic in which side-task stimulus presentation depends upon concurrent performance on the pri- mary task has also been investigated (Damos, l972; Damos and Roscoe, l970). It also works, but not so well as the simple self-adaptive task just described. The measurement of residual attention in a standardized manner under specified flight situations, whether in actual or simulated flight, offers a promising common basis for establishing the workload demand and blunder proneness of area navigation, vertical guidance, and other types of flight directing and control systems. As such, it represents a potential method of demonstrating compliance with objective performance standards for airborne equipment certification. Stanley N. Roscoe Aviation Research Laboratory University of Illinois

48 INTERFACE MAN-COMPUTER DATA ENTRY I. Scope The purpose of this study is to assess long range needs in the interface between the internal digital manipulation of the electronic signal and the outside world. For want of a better terminology, we shall call the communication which crosses this interface "data". Such data is generated by people, physical processes, transaction documents, pictures and miscellaneous other source documents. As well as the physical transduction of the data itself, the study is con- cerned with human factors, formatting, and programming support issues which facilitate the generation of entry of data. The data entry area is not a cohesive body of technology. More importantly, the value of a technology is largely deter- mined by the system and operational environment within which it resides. Thus, there is no attempt to draw a sharp defini- tion of this area. Rather, this report attempts to cover the interesting approaches of physical data entry and a look into problems of data generation that the user faces in dealing with the hardware. The latter subject leads into a large number of programming and human factors issues. In the long run, these will become the dominant considerations in design- ing an entire man-computer system. II. Perspective Historically, the major impedance to the growth of the computing business has moved outward. Until the present era the cost of hardware, especially online storage, was a major factor in the customer's decision to put an application on a computer. With present computers, the complexity and stability of the software system have become the first consideration in a new application, or the expansion of an old one. The present development emphasis is to clean up the system structure and make it much more permissive of application expansion and the programming of new work. It seems reasonable then that growth in function and facilities will be strongly dependent upon our ability to interface the computer with an enormously expanding variety of day-to-day endeavor. Consequently, our ability to insert the computer transactions in a natural way must be a major focus of research effort. Whenever a person is directly involved, data output is an integral part of the data entry functions. For example, conversational interactive data entry requires some degree of data display. The system must enable the person who is monitoring automatic data entry or executing manual data entry

49 to supply missing data or to re-execute part of the entry process. To the extent that intervention is required by the nature of the data entry task or by the nature of the system, the monitor/enterer must be provided with a context in which he can determine the occasion for and degree of exception-handling needed. The system must either be con- figured to recognize the error, or display a subset of data which enables the user to recognize it. This is one of the factors which makes the successful configuration of entry and display techniques application-dependent. III. Trends There are a number of major trends which could dominate the theme of data entry for the period of interest. These are: A. The economics and logistics of handling data several times before it gets into the machine yields strong emphasis on capturing data as near its source as possible. While this direction seems to pervade all applications, there are two principal keys to its rate of progress. First, as the data capture hardware gets further from the central system, its duty factor of usage tends to decline. Consequently, the absolute cost of the capture device must go down correspondingly. Secondly, even in instances where the duty factor of usage is high, such as in gas station credit card transactions, the cost of communication to the central system becomes a major com- ponent of over-all transaction cost. While many applications will allow remote buffering of information in machine readable form, it is highly desirable in most financial transactions to have two-way communication with a computer in order to obtain positive authorization for the transaction and to mini- mize float. B. The fastest growing segment of data entry is in the consumer area. Here the principal requirement is to read a credit card, identify a piece of merchandise, or accept a simple query. The problem is to prestructure the transactions in such a way that the general public's interaction with the computer is extremely simple. C. Optical character recognition (OCR) will continue to grow as an input mechanism, especially where reasonable volume is involved. The principal trend in OCR equipment is to reduce the threshold cost and to reduce the constraints on its usage, principally in the area of alpha-numeric hand printing. IV. Summary l. The major hardware needs in the data entry area are for better terminal printers and large screen displays. Both of these components are currently major cost problems.

50 2. Optical character recognition will grow in importance in the future. Technical advances are needed here to permit a broader expansion of data entry capability to hand prepared documents. 3. Improved data entry technology is desirable in appli- cations such as retail check-out systems. 4. Human factor studies and improved speech recognition capability is required to expand voice communication capa- bility with computers. 5. Other than the need for better graphic display, the principal need in graphical data entry is for better human factored software. 6. At the man-machine interface, we understand the machine much better than the man. There are both hardware and software design issues, but software issues are dominant and growing. We have almost no predictive design technology in this area. 7. For purposes of security and authorization, systematic studies of ways to-positively identify a person at a terminal are desirable. 8. A cursory investigation of physiological phenomena yielded no practical data entry mechanisms. There are more scientifically oriented investigations that might bear fruit later. 9. "Data Entry" is not a cohesive body of technology on system function. It should be viewed as part of the organic whole of an application. There always are several alternative data entry technologies for the same application and their relative value can be determined only in the context of the entire application environment. V. Technology Issues A, Key Entry Stand alone keypunches, clustered key entry stations and terminals will be included in this category. The trend is away from the stand alone units to clustered and on-line entry. Key entry will remain an important computer interface for the foreseeable future. The trend in OCR is toward lower cost and fewer constraints on the input document. Thus, the impact of OCR on key entry will continue to increase, but not enough to remove key entry as the dominant data entry mechanism.

51 Keyboard design today lacks discipline. Not only are many mechanical aspects still open to investigation, but the logical structure seems to require new study each time a new application is considered. The following list outlines some of these issues. - Size of keys - Key spacing - Operating force - Displacement - Keyboard shape (Concavity) - Keyboard slope - Feedback (kinesthetic, auditory, visual) - Keyboard format - Key interlocks - Application sensitivity B. Character Recognition l. State of the Technology Machine printed and handwritten documents or pages are pervasive in our society, and OCR is a candidate for data- entry whenever information from documents is now done by manual keying at a cost (mostly labor) which approaches one- fourth the total cost in contemporary DP (data processing) installations. Replacement of a major part of this labor with labor saving OCR equipment is desirable. There are other alterna- tives to OCR, such as placing key entry terminals at the source of data, or recording data in machine readable (coded) form at the source. OCR is likely to have a significant role as long as paper documents are a commonplace medium for stor- ing and transporting information. To date OCR has not had wide acceptance. The reasons are: l. The machines are expensive. 2. Installation costs are high. 3. Inputs are relatively constrained and on special forms. 4. Potential users have not been adequately aware of the benefits of OCR. Although the rate of growth of OCR is uncertain, there is strong intrinsic motivation for OCR as a direct and obvious method of capturing data at the source. Growth of OCR is limited because we haven't adequately solved the problems associated with OCR, not because the potential application area is small. It is probable that certain mild constraints on the input will be necessary and acceptable for some time in the future.

52 The problem of machine recognition of cursive hand- writing has been recognized as very difficult - in a league with complex pattern recognition and speech recognition. Work is in an exploratory stage. There is need for innovative and imaginative approaches to make progress in this area. 2. Technology Problems in OCR Solid-state scanners are now feasible, and reductions in memory and logic costs allow a shift toward programmable machines rather than hard-wiring. a. Current document transports for OCR and MICR (magnetic ink character recognition) handle a wide variety of documents at high speed remarkably well. There do not appear to be specific technology barriers in this area, but we can assume that cost would be a problem in any system aiming toward the low-cost end of the market. b. Current scanner approaches (LED or Photo-FET arrays) appear quite good. c. With relatively unconstrained input, finding the desired field of characters is a major problem. Algorithms are needed for performing this function in memory. d. Measurement design is still largely a manual process without formal guiding principles. Measurement selection and design of decision functions is computer-aided, but the pro- cedures don't have a strong theoretical foundation. e. Post-processing to apply contextual rules to fill in characters that were rejected is currently an application dependent art. General principles or procedures would be desirable, but may not exist. f. The low-volume user may best be served by a scanner terminal with centralized recognition. With telephone lines, some form of compaction is required to get enough throughput. It might be desirable to do recognition on the data in com- pacted form if properly coded. Another alternative is to push toward broad-band transmission in the first place. C. Credit Cards While the embossed card will remain for a long period of time, magnetic stripe credit cards could offer significant advantages. This approach allows the encoding of a great deal more information on the card and additionally allows the system to record back credit or other status information.

53 In addition to providing more user data on the card, the need for security against theft and inadequate credit balances causes a trend toward online systems. The principal thrust seems to be the use of the credit card in connection with a memorized four-digit personal code. Some systems have additionally gone to cryptographic approaches of coding the data before transmission across telephone lines in order to further protect the personal codes. While it is believed that the four-digit code and credit card serial number will be sufficient to drastically reduce the current losses, this may not yield a sufficient identi- fication for some transactions. Therefore, it is worthwhile to consider more positive, but also simple means of additional personal identification. D. Speech Recognition Speech recognition has the potential for greatly enhancing the man-machine interface. Examples would be giving an astro- naut additional control capability when his hands are occupied, or giving a mail clerk the capability of voice control to supplement manual control of a mail sorting system. Speech control successfully implemented could supplement, and in some cases, replace the keyboard as an input device. l. Classification of Speech Recognition Problems There are four basic dimensions to the speech recognition problem: the range of speaker variability; the complexity of the material spoken; the background noise level; and the channel distortion between the speaker and the A/D converter. Constraints placed on each of these dimensions give rise to a classification of speech recognition problems and speech recognition capability. The basic alternatives as concern speaker variation are single speaker and multiple speaker. Multiple speaker breaks down by dialect, by sex (women have proven more difficult than men), and by age (particularly children). Thus, we can talk about the capability to recognize the speech of a single dia- lect, of all female Speakers, of any American dialect, etc. The basic alternatives as concern the complexity of the material spoken are isolated words and continuous speech. The problem of recognizing isolated words is called word recog- nition. Word recognition capability is roughly measured by the number of words recognized. Continuous speech capability depends upon the vocabulary size and upon the language. It varies all the way from continuous sequences of a few words, through artificial languages with limited vocabularies and a simple syntax, to natural English. The background noise level varies from essentially zero to a factory environment.

54 The channel distortion practically breaks down into two alternatives: input over the telephone, or over a higher fidelity channel. 2. The Difficulty of the General Speech Recognition Problem The general problem of the machine recognition of the spoken vernacular of an unknown American speaker is a problem of immense complexity. It is a pattern recognition and linguistic processing problem par excellence. The acoustic correlates of the basic phonetic elements of speech vary dramatically; with dialect, with speaker, with linguistic context, and with emotion. Even at the word level for a single speaker there is great variability. There is strong co-articulation between words, and the total acoustic correlate of a word depends strongly upon sentence context. There are no word boundaries except at pauses. Further, syntactic and semantic "knowledge" sufficient for the understanding of utterances in a language is a requirement for accurate human phonetic transcription of that language. People only speak carefully enough to be understood by other people, which means that any general speech recognition device must be able to capitalize on the redundancy in speech and hence cannot operate as an acoustic processor alone, i.e., it must do sophisticated syntactic and semantic processing. The achievement of this acoustic and linguistic pro- cessing function in an effective way short of the sensing and processing capabilities, as well as experience of a human being, is a major research challenge. 3. State-of-the Art a. Word Recognition Present capability for word recognition for multiple male speakers is twenty to fifty words with 95% accuracy. This has been achieved in a number of places. This capability has not changed substantially during the last ten years. The problem is solved by applying various standard pattern recognition techniques to the entire word. It is not solved in any fashion which is likely to be extendable to continuous speech. Present capability for word recognition for a single speaker is fifty to one hundred words with 95% accuracy. This has also been achieved in a number of places and is done in the same fashion as for multiple speakers. In both cases

55 the performance is somewhat less over the telephone. This one hundred word capability can be compared with the l0,000 or more word recognition capability of the human being. One company, Scope Inc., markets a word recognizer. It is a single speaker device and recognizes l0 to 54 words. The only other system in actual use to our knowledge is the IBM Raleigh FE system. It has a multiple speaker capability which can recognize l3 words from a telephone input. Message accuracy is measured by the discrimination between yes and no which is done with 98% accuracy. The thirteen words are recognized with 95% accuracy. b. Continuous Speech Recognition The system put together at IBM Raleigh with Air Force support is the only non-trivial complete continuous speech recognition system. It operates on a 250 word vocabulary with a simple, fixed syntax language spoken by multiple male speakers. The 250 words are the most commonly occurring 250 words in English text. The system has processed 34 sentences and recognized 9 of these perfectly. It can recognize 75% of the words, and, depending upon the class, 50% to 70% of the phonemes in these 34 sentences. This work is noteworthy because it is the first time a complete system (with both acoustic and linguistic processing) for continuous speech has been assembled, and because it gives surprisingly good performance for an initial attempt. A system put together by Reddy at Stanford operates as a limited continuous speech recognizer. It recognizes commands to the Stanford block stacking programs. Both of these programs (The Raleigh and the Stanford) would require about 5 MIPS* to operate in real time on a general purpose machine with no special hardware of any kind. The Raleigh system includes context-dependent phonetic segmentation and recognition and linguistic processing. These two ingredi- ents were recognized as essential, in the intervening period, to any successful program. 4. Nature of Speech Recognition Research to Date Research on the automatic recognition of speech began over fifteen years ago. There have been a number of efforts including Bell Laboratories, RCA, Litton, Lincoln Laboratories, IBM, and others. The initial activity (l955-60) was concerned with continuous speech recognition and was a slow process of disillusionment because of the general difficulty of the problem. *Millions of instructions per second

56 Subsequent projects were mostly concerned with word recog- nition and resulted in the widespread uniform level of achievement cited above. Up until the present ARPA program discussed below, there have been no large scale federally financed efforts in speech recognition. In addition to the activity in industrial and defense laboratories, there has been continuing work at universities and a substantial number of graduate theses. There has also been considerable research activity in linguistics, speech science, and pattern recognition, which is pertinent to speech recognition. 5. The ARPA Program ARPA has undertaken a five year program in continuous speech recognition at a projected cost of approximately 15 million dollars. Ten million of this is a re-orientation of old costs, and 5 million represents new expenditures. This effort follows the recommendation of a study committee which met during the Summer of 1970 and which produced a report "Speech Understanding Systems", May l97l, which is available from the National Technical Information Service, Department of Commerce, Springfield, Virginia. The goal of the ARPA effort is a system for the recog- nition of continuous speech from multiple speakers in a quiet room with high fidelity input. The language would be a command language, generally not Englishlike, with a highly constrained syntax and a vocabulary of approximately l000 words. ARPA is interested in a com- plete system which integrates speech recognition into the solution of a reasonable and well specified task. The goal is a demonstrable system by l976. There will be a number of parallel efforts over the first two years. The major groups involved are SDC, SRI, BBN, Lincoln Laboratories, and Carnegie-Mellon. The Speech Communications Laboratory at Santa Barbara, the University of Michigan, and Univac in Minneapolis are also involved; the latter under a subcontract with BBN. At the end of two years, if the results are sufficiently promising, one or more of these contractors will be selected for a subsequent three year effort yielding a complete system with viable applications. Other features of this program are: a steering committee to coordinate the effort; a summer institute to promote tech- nical interchange among the individual projects; use of the ARPA network to facilitate sharing of hardware, software and data; a common data bank (maintained by the Lincoln Laboratory)

57 to insure a valid comparision of results; and continued financing of university research. 6. The MITI Program in Japan The Japanese government has embarked on a l00 million dollar joint government-industry program for a "Pattern Information Processing System" (PIPS). This program is directed toward character recognition, two-dimensional image processing, three-dimensional image processing, and speech recognition. Their present goal in speech recognition is the recognition of l00 isolated words. However, the detailed planning of the speech effort does not begin until next year and this goal will likely be modified. There has been no decision as to the amount of money which will be spent on speech. The plan is to have a pilot speech recognition system in five years. 7. References The following is a brief list of references for obtain- ing general background in the speech recognition area. "Speech Understanding Systems", A. Newell et al, May l97l, ARPA Report, National Technical Information Service, Spring- field, Virginia. "Approaches to the Machine Recognition of Conversational Speech", K. W. Otten, Advances in Computers, l97l, pp. l27-l63. "Man Machine Interaction Using Speech", D. R. Hill, Advances in Computers, l97l, pp. l75-230. "Automatic Speech Recognition: Literature, Survey and Discussion", S. R. Hyde, l968, Research Dept., Report No. 35, P. 0. Research Dept., Dollis Hill, London, N.W. 2. "Machine Recognition of Human Language", N. Lindgren, IEEE Spectrum, No. 7, l968. "Whither Speech Recognition," J. R. Pierce, Journal of the Acoustical Society of America, l969, pp. l049-l05l. E. Image Entry Images, two-dimensional or three-dimensional signals, constitute a significant class of data. There exist enormous quantities of information in the form of visual images which are now processed manually. The automatic or semi-automatic acquisition and processing of this information could substan- tially reduce this manual task. Generally the processing of images is complex, and to date, except for the simplest tasks, computers have not been effective and efficient. However,

58 it can be expected that the rising cost of labor, shortage of trained personnel, the decreasing cost of computation and memory, progress in methodology, sensors and effective man- machine interaction would bring about a major penetration of the computer into these areas. 1. Data Types To provide a perspective, a list of data types for entry to the computer other than the keyboard entry is given in the following: a. Time Signals l. Voice Speech recognition Speaker verification and identification 2. Seismic Discrimination (nuclear explosions, earthquakes) Exploration 3. Biomedical signals Electrocardiograms Electroencephalograms Etc. 4. Sonar and radar b. Two-dimensional patterns l. Characters Stylized fonts Mixed fonts Hand printed characters Cursive writing On-line character recognition 2. Fingerprints 3. Maps 4. Text Format recognition combines A/N and line drawing 5. Bubble and spark chamber photographs 6. Aerial photographs/satellite images Clouds/weather Crops Targets, reconnaissance Pollution control Earth resources

59 7. Electron micrographs Material and integrated circuits Biological specimens 8. Biomedical images Scintigrams Chromosome Bacteria Pharmaceutical Smears Cells and Tissues Brain Sections X-ray films c. Three-dimensional patterns Scene analysis Robotics/artificial intelligence Manufacturing The types of data in the first part of the list are either one-dimensional or intrinsically binary (black and white) patterns, and only those on the second portion of the list are images with grey scale. 2. Environment There is growing interest and activity in pattern information processing, not only in USA, but in other coun- tries, and notably in Japan. The Japanese Government decided in l97l to invest over $l00 million for a period of eight years in research and development of a "Pattern Information Processing System" under the National Research and Develop- ment Program, with active industry and university participation. Indicative of the prevailing general interest in the U.S., the National Science Foundation held a conference in February l972 to discuss research opportunities on "pattern information processing." However, as yet there is little commercial impetus to develop image entry and/or processing capability. This apparent lack of commercial interest can be, in part, attri- buted to the fact that almost all successful computer appli- cations have so far been in the areas of business data processing, management information and scientific computation, and that the major sources of images are in medicine, science and applied technology. The impetus and motivation for acquisition and processing aerial photographs and satellite images are being provided

60 by the government. A few major image processing centers will be established and tasks performed on national and state levels. There are over 40 known groups at universities, medical centers, government laboratories, independent research institutes and industry laboratories which are engaged in computer processing of biomedical images, electron micrographs and 3-d patterns. The efforts range from academic inquiry, research, development, and feasibility studies to clinical implementation. 3. Needs a. Image entry b. Image processing In order to make computers effective and efficient in handling visual images, it is necessary to provide adequate means for image acquisition and display. But this is not enough; one has to provide image processing methodology or programs. This is especially true in medical applications. It is not realistic to expect practicing medical doctors to develop their own processing algorithms and programs. As a group, medical doctors are skeptical about the usefulness of computers in their profession other than for keeping track of patient information and bills. In order to penetrate successfully into the handling of biomedical images, it will be necessary to develop basic processing software, as well as to solve the problem of image entry. This dependency upon development of processing techniques and algorithms is a key difference between image entry and other types of data entry, say character recognition. In the latter, once the characters are recognized, the processing of this information is well defined. Unfortunately, it's generally not the case in handling images. Merely to get images in and out of the computer is not sufficient; how to process them in many cases is not understood. The user may say "enhance the picture" or "find the ventricle"; how to achieve this is usually not a trivial task. F. Graphical Data Entry Graphic Data Entry is the creation and maintenance of a coded graphic data base representing both the image and the inherent structure of the information on documents containing combined alphanumeric and graphic data. The emphasis here is not on highly interactive graphic design, but rather, on the efficient entry of large volumes of graphic data.

61 Unlike the image processing problem, graphical data admits to an explicit, albeit complex, relationship between the computer representation of the picture and the semantics of the object. Thus, the issues of recognition and mani- pulation are enormously simpler. Correspondingly, the costs associated with processing and storage cease to be dominant. The focus thus moves to the entry interface. At this point, the dominant aspect is the human factors of the programming which supports the creation and entry of the data. Secondarily, the entry hardware is an issue. Applications fall into two categories: (l) Mapping - The processing of documents which represent geographic boun- daries or areas and the location of facilities relative to this geography; and (2) Drawings - Combined alphanumeric and graphic information usually relating to the production (or building) and maintenance of products, processes or procedures. A list of typical applications follows: - Mapping and cartography - Computer-aided drafting - Page layout and composition - LSI Chip design - Numerical machine control - Electrical circuit diagram - Piping layouts - Plant floor plans - Pert charts - Truss design - Chemical structure diagrams - Architectural drawings - Program flowcharts and documentation The data base generated as a result of graphic data entry is the key to the success of these applications. It is used for two purposes: (l) Computer-produced documents at a level current with that of the coded data base; and (2) Pro- cessing and manipulation, by application programs, of informa- tion represented on the documents. The principal value of such systems is to increase greatly the productivity of the people involved in design and drafting. Thus, it is necessary that the device "feel right" to people skilled in these fields. In the software area, the major need is for a customizer which will allow the non-programmer to fit the system to his particular needs. This would include special symbols, drafting conventions, etc., as well as the facility to incorporate pro- cessing macros particular to the application. ,1

62 G. Human Sciences There seems to be little doubt that the focus of com- mercial and technical interest is moving outward from the Central Processing Unit (CPU) toward the customer's work and the user himself. As we make this movement outward, more uncertainty is encountered in the product strategy and design process. Not only is it more difficult to discern what is needed, but it is much more difficult to determine if a parti- cular design meets the stated requirement. It is easy to compare arithmetic units, but more difficult to compare information systems. There is no technology to allow the comparison of conversational languages. A cursory look at our scientific and technical resources shows why we understand much more about what goes on inside a semiconductor device than we do about what goes on in the mind of a user as he attempts to get his work done in a computer. The latter issue is becoming more important than the former. From the perspective of data entry, work is needed to develop a design and evaluation methodology for those com- ponents of hardware and software through which data is entered. Classically, such work has been called human factors or be- havioral science. While this work is heavily dependent on those disciplines, there appear to be a number of areas which have not been confronted by these fields, but are of growing importance. The following issues define the scope of some of these. l. Issues a. Technology Evaluation and Design As we develop new I/O technology components, a comparable amount of effort is needed to understand and synthesize the man-machine systems within which they reside. The value of the technology can be established only in that context. Current technology, itself, is not well understood from a human factors viewpoint. Other than the aspect of noise associated with the typewriter terminal, there is little agree- ment on whether hard copy or Cathode Ray Tube (CRT) display is better for conversational computing. Perhaps this issue is unresolvable, but work is required to understand what factors make it so. b. Human Factors of Programming Languages As we proceed to design new languages which will facili- tate more users, there is no technology which will predict

63 how people will like them. Which commands, for example, correspond naturally to the way people want to think about their problem and which are there because the system pro- grammer put them there to make his job easier. Field studies show that one of the major costs in application development is in the area of debug and checkout. It has further been shown that the bulk of errors are errors in program structure (as contrasted to coding) and application concept. A major issue is the design of effective languages that will encourage the layman to bring more of his data pro- cessing work into the system. The intent here is to eliminate the application programmer from a broad class of simple trans- actions and at the same time facilitate the user in his task of data generation. A striking example of a system which facilitates data entry is the display oriented project led by Dr. Douglas Engelbart at Stanford Research Institute. Here the users are expected to undergo a great deal of training. The payoff is an extremely powerful data and text manipulation capability with which the information can be sorted, indexed, retrieved, cross-referenced, structured, etc. The system also provides excellent capability for com- munication among users. The key to the power of the system and also the source of the training need is an extensive use of context. A great deal of attention is paid to anticipating what the user might do next. By taking account of the context of the user's current state, extremely brief (one or two characters) and powerful commands are the most commonly used. A very interesting issue raised here is the tradeoff between the user's throughput obtained from an extensive support structure versus the training required to make use of it. There does not seem to be much understood about this issue as well as the larger issue of why there is such a large variability among the productivity of programmers. Systems such as these are invented rather than synthesized from funda- mental principles. They are successful to the degree that they align with the way the user wants to think about his work. R. A. Henle G. C. Bacon P. D. Welch IBM, Hopewell Junction, New York

64 INTERACTIVE CABLE TELEVISION SYSTEMS Over the past twenty years cable television systems in the U.S. have grown from l4,000 home subscribers in l952 to a total of about 7 million as of the end of l972. Current subscriber growth is at a 20% annual rate. From their early function to provide TV reception in remote areas not reached by over-the-air television, cable television applications have moved rapidly to higher quality color reception, to greater availability and choice of pro- grams, and to local programming. Additional applications have come under consideration as it has become apparent that this significant growth of cable systems will enable them to reach large numbers of subscribers with many channels for use in education - training, health care, specialized informa- tion, shopping, premium movies or other entertainment at an additional fee, and other specialized services. It has become evident that many more applications are possible, but that they would necessitate a two-way system either to collect automatically organized information on a home terminal, to enable the home terminal to interrogate information centers, or to answer questions. The following list illustrates some of the possibilities: In addition to the usual line-up of cablecasting news and weather, CATV operators will be able to offer the broad- band "wired city" such new two-way services as: For the home - burglar and fire alarm service; banking; shopping; elec- tronic newspapers; electronic mail; special interest programs on order; opinion polling; and meter reading. For business - electronic mail; data retrieval; computer time-sharing; document transmittal; video conferencing; market testing; credit card validation; and facility security. For education - computer-aided instruction; data retrieval; computer time sharing; and access to centralized library services. For government - video conferencing; training courses; computer access; fire and burglar detection; picture and fingerprint record retrieval; remote "line-up" of suspects; and automated traffic control.

65 Some of these services can also be provided by the telephone system. It is not yet clear which communication system will be found most convenient in the future for various of these present and future applications. It may be that the telephone system can best take care of some services and the cable television systems can most effec- tively handle others. Because of their great capacity and flexibility, cable systems, which will develop substan- tially on the basis of a few major applications, will have the capacity for many additional ones. In order to open the door to all these possible applications, many systems are already being planned and built with upstream capabilities in the form of an addi- tional cable, or by reserving a part of the bandwidth in the single cable for the return signals. The additional terminal equipments, at the home or at the cable system headend can be installed later as they are required. Several such interactive cable systems are in being, i.e., New York City; Spartanburg, South Carolina; Orlando, Florida; Irving, Texas; and El Segundo, California. Some l,000 FCC authorization certificates have been issued through April l973 for additional systems with this capa- bility. Future FCC rulings will determine many of the cable developments and may have to settle the conflicts between the use of the narrow bandwidth (telephone) and the broadband cable systems. Thus, broader bandwidth on the home telephone system will help to stimulate two-way video phone applications which may well compete with some of those permitted by the cable. Many studies of these problems have been made by foundations, industry, the FCC, and much has been published already. A few illustrative references are attached. There is no point in covering extensively in this report the contents of all these studies. However, a number of preliminary conclusions will be suggested, keep- ing in mind that our interest is in attempting to identify some areas where research could be pursued or expanded to benefit additional broadband applications. While some of the two-way applications using signals raise no social questions other than public acceptance and reasonable cost, this is not the case of other applications which raise the spectre of possible undue influence and a lack of balance in the presentation of issues and in the sampling of public opinion.

66 While with a limited number of channels reaching all the citizens of the country it seems possible to have at least a degree of balance in the use of the media by politicians and propagandists of ideas because of the clear and large visibility of what happens, the avail- ability of l0 times as many channels will expand this visibility considerably. It will enable those political candidates who are less well off financially to afford to reach an audience; likewise, the political candidate can focus his message on a specific audience for maximum impact and even determine audience reaction. If the upstream interactive capability is used to get "instantaneous" reactions to ideas and issues pre- sented on the cable system, there is a grave danger that the opportunity for a consideration of balanced viewpoints will not exist. The instantaneous reactions may represent a maximum of emotion and a minimum of logical considered response. Time is necessary to think non-emotionally on sensitive topics. This issue is complex,as it may not be practical to present a subject on television on one day and collect opinions and responses one or two days later. How can we insure that the various sides of a complex question have been well presented before opinions are polled? Illustrative of this high variability of public opinion by polling are the wide swings in popularity ratings of politicians involved in current highly charged issues. A factor of interest is that at least the hysterical emotional involvement often created in mass meetings is practically avoided by the physical separation of the audience and the distribution of information to home audiences of limited size. The question is whether there is some useful research to be done in this aspect of telecommunications which could result in beneficial ways of using interactive broad- band systems, beneficial in the sense that the impact of imbalanced viewpoints and emotional responses would be lessened. In this connection, studies might usefully be initi- ated on the dangers of excessive communications and on the heightening of conflict by over-communicating an issue. Let us assume that whatever anyone thinks at any one time would be available to his associates. While many misunderstandings could potentially be eliminated, in fact, life could become impossible as there are many thoughts, ideas and plans which have to be filtered, reconsidered, or

6-7 suppressed, even between persons related by strong love or friendship. As long as persons use their minds as they are meant by nature to function, there are limits to the amount of information which can be exchanged and absorbed usefully. These limits are impossible to define clearly as they will vary enormously with time and circumstances. However, studies of understanding and misunderstanding in communi- cations, and of the practical limitations on the flow and absorption of information to allow social groups to coexist peacefully may be proper in view of the development of rapid, high capacity and interactive communications. Another danger involved in interactive communications is its inherent potential for the invasion of the viewer's privacy. The same technology that permits the cable tele- vision viewer to interact with the media, could enable the politician to determine which citizens are watching his program and which are watching his rival. Of course, education, training, telemedicine, informa- tion access possibilities, and many other interactive functions are expected to be very beneficial. The high capacity and responsiveness of this technology will call for the practical education of our citizens to its great potential in enabling them to interact with the media and also to its dangers. Abuses of the technology must be guarded against by an educated, concerned and alert public. As for the scientific aspect of research in broadband systems, the following items are presented of possible efforts to be pursued or expanded on: 1. The apparently great potential capacity (10^ bits/ second) and flexibility of optical communications systems which are already under development. 2. Studies of the means of collecting detailed comments on an issue. 3. The present upstream signal systems in which commu- nication with a large number of terminals is effected, is subject to noise, and generalized disturbance. Better solu- tions are needed based on research in combination with item 2 above.

68 4. It would seem that as soon as the audience in an interactive cable system passes a hundred, effective audience participation becomes practically impossible; since the audience could reach thousands or even hundreds of thousands of persons, the problem of audience partici- pation is not readily solved. It is possible to conceive of much improved technical solutions to permit the trans- mission of many comments per person and their correct analysis, without ignoring the importance of the non- response in the assessments. 5. At the moment the U.S. seems to be leading the world in broadband systems, although Japan is moving rapidly in this field. However, the present systems use elementary technologies. This was probably an advantage at the start, but may prove to introduce system limitations later. Relatively little research and development is being done in this field, and while the U.S. is ahead in applica- tions, some advanced work is well justified to keep the lead in technologies for the applications of l0 - 20 years in the future. Possible subjects for research are as follows: a. Broadband cable network research. The publications in this area have been conceptual or social science-oriented. An investigation needs to be made into whether research in devices and amplifiers has proceeded sufficiently to enable economic implementation of the broadband communications network. An investigative effort is proposed looking at solid state device research in appropriate centers of knowledge and in analyzing amplifier concepts of specific manufacturers. b. Interactive home terminal research. This intriguing concept is too costly today. An investigative effort is proposed into the principal components (display, keyboard, etc.) to determine whether research should be directed towards achieving elements which could have the potential for achieving lower cost terminals. H. Busignies ITT New York City

69 INTERACTIVE CABLE TELEVISION SYSTEMS References Cable Television: A Handbook for Decisionmaking, Walter S. Baer, The Rand Corporation, Santa Monica, Calif., l973. Campaigning on Cable Television, National Cable Television Association, l972. National Cable Television Association, Inc., Statistics on Cable Television Systems. On The Cable: The Television of Abundance, Report of the Sloan Commission on Cable Communications, McGraw- Hill, l971. Project Minerva, Preliminary Findings of Electronic Town Hall, Center for Policy Research, Inc., New York, N.Y., l973. Telecommunications: Its Impact on Business, E. Bryan Carne, Harvard Business Review, July-August, l972. Two-way Cable TV Systems, Ronald K. Jurgen, IEEE Spectrum, November, l97l.

70 TRAINING AND EDUCATION OF PROFESSIONAL TELECOMMUNICATIONS ENGINEERS I. Introduction and Problem The telecommunications industry, along with its supporting science and technology, has grown very rapidly in the past decades. It represents a very substantial part of the US and_ world economy. It has a vital influence on other industries and businesses that represent even a larger segment of the economy. Furthermore, telecommuni- cations is vital from national security and defense standpoints. The problem we wish to discuss is the education and training of professional calibre telecommunications engineers to meet present and projected needs of the industry and the country. We would like to know whether in fact there are appropriate curricula and academic courses for specializing in telecommunications engineering, whether these courses and curricula are fully appropriate to the present and future needs, insofar as we can anticipate them. Relevant to the above is the observation of some of us in the industry, that in spite of the recent apparent abun- dance of engineering talent available in the electronics industry due to the depressed state of the defense and aero- space industries, competent professional level telecommuni- cations engineers available to assume responsibility are really very scarce. Many professional level jobs are, therefore, filled by marginally qualified people. II. Questions The problem can best be highlighted by a series of relevant questions. How has the need for professional manpower been met during its large growth period between World War II and the present? How is professional training provided now? How does industry meet its needs? A. Related to Schools: From the standpoint of the schools and universities, are there the proper curricula for telecommunications engineers? Is most of the training actually for electrical engineers and electronic specialists, and, if so, is this the optimum way to produce telecommunications engineers of good professional calibre? Are there enough graduates for the present needs? Are there likely to be enough for future needs? Is there proper training for research engineers whose main professional

71 objective is to advance the state of the art? Is it practical to provide a proper professional level telecommuni- cations training in a four year undergraduate course? Is there not need for a course tailor-made for the telecommuni- cations profession - including suitable graduate curricula for specialties within telecommunications, i.e., common carrier, broadcast, international telecommunications, data networks, computer technology, etc.? B. Related to Industry: Does the training include appropriate discussion and understanding of the characteris- tics and problems of the telecommunications industry; how is it organized, managed, regulated? What are the competitive aspects and international factors? How is it affected by the general economy? How is it influenced by the Department of Defense needs, or other government needs, whether federal, state or local? C. Related to Society: What about the interface between telecommunications and society? What are the signi- ficant economic and social aspects like cost and price policy, monopoly versus competition, problems of regulation, privacy, and the fairness doctrine? What is the real and potential impact of telecommunications, and more specifically, CATV and other broadband communications and data networks on education, health services, management of welfare and main- tenance of public safety? Are there suitable curricula or centers of study for mature graduate students, pre and post doctoral, to study, understand and elucidate the really significant issues for the interested public so that it will be possible to evolve sound long range policies, free from undue political and industrial pressures and conflicts? III. Professional Training - Present and Recent Past How is professional engineering training in telecommuni- cations provided at the present time, and how has it been provided in the recent past? There have been a number of avenues. The most straightforward has been by formal training as an electrical engineer with an undergraduate degree, followed by on-the-job training in industry or government. Sometimes the undergraduate work is followed by one or more years of specialized training possibly leading to a master's or doctor's degree. This, in turn, leads to a responsible job in some area of research in telecommunications or electronics. Sometimes the formal training is in computer sciences, physics, mathematics, or other related areas of science. This then provides an excellent foundation for research in tele- communications .

72 Sometimes in the past, and particularly during the World War II period, one got a start in telecommunications with little if any professional training, but rather as an amateur radio enthusiast with practical electronic experience. This, coupled with a number of years of significant on-the- job training in industry or in the military services, has led to successful careers in telecommunications. Many of this group are now reaching retirement and need to be replaced. However, telecommunications has not been established as a bona fide engineering discipline and profession on its own. The courses and curricula have been fine for training students in specialized fields of research vital to telecommunications. Specialists have been trained in amplifiers, detectors, antennas, computers, memories, integrated circuits, electromagnetic theory, information theory, etc. What has been neglected is the train- ing of the well-rounded sound professional engineers who under- stand the basic technology, and the problems of the industry and the profession, as well as their impact on other industries and society. The latter capability, if acquired, is through on-the- job training and experience. It has, in fact, been a good way; but is it optimum for our present and future needs? With some revision and modification of the courses and curricula in the schools and universities a sounder pro- fessional training in telecommunications can be provided. This will be advantageous both to the engineering profession and to the telecommunications industry. Many schools and universities are beginning to appreci- ate that the courses offered are no longer optimum from the standpoint of meeting the present and projected needs of the telecommunications industry. They are beginning to revise their programs and introduce some innovations in this field. Among these are MIT, University of Illinois, University of Colorado, George Washington University, Brooklyn Polytechnic Institute, and Purdue. The backing and guidance of the National Science Founda- tion and the National Academy of Engineering would help this healthy tendency. IV. Suggested New Elective Courses for the Telecommunications Engineering Curriculum ~" What are the items of professional training in tele- communications that are generally missing? Following are some examples: l. Telephony and common carrier industry character- istics and problems.

73 2. Network and network planning. 3 . Transmission systems 4. Switching systems. 5. Traffic engineering, statistics of communications, its measurement and analysis. 6. Electronics technology's impact on common carriers and future telecommunications. 7. Defense communications - special needs and problems, security. 8. Telecommunications system architecture and system engineering. 9. Economics of telecommunications, monopoly, competi- tion, cost, price, etc. l0. Specialized carriers. ll. Data and information systems - need for standards. l2. Computer technology and its impact on telecommuni- cations systems. l3. Radio spectrum - usage and compatibility. l4. Regulation and the legal issues in telecommunications. 15. Broadcasting - television and radio - special issues and problems. 16. CATV. l7. Social impact of telecommunications. l8. Research trends - future outlook, l9. International telecommunications - satellites, undersea cables, international trade. 20. Telecommunications application problems. Health Navigation Law Enforcement Special state and local issues Regulation Surveillance - radar Education The list is obviously not complete, but is representa- tive of the types of studies which professional training must include.

74 It does not follow that every professional telecommuni- cations engineer must be an expert in each of the above areas. However, a professional calibre engineer with the proper training should at least be aware of most of the fundamentals of these problems and be reasonably expert in a few specific areas. A good research engineer, who is interested primarily in advancing the state of the art in a specialized area, need not necessarily concern himself with any of the above. But a good telecommunications professional must have the broader concern, because his own welfare, that of his pro- fession and of the industry depend on it. Thus what is being advocated is not "a drastic change", but a significant reorientation for the telecommunications professional. It is important to appreciate that society's needs and priorities are changing, as is industry and competition from abroad. The educational system ought to take this into con- sideration. There is a tendency to be complacent and pat ourselves on the back because we have developed an outstanding technology in the past three decades which was very effective in a war, cold war, and "Sputnik" environment, but without modification is no longer as relevant to our needs as it once was. V. The Need for Telecommunications Professionals It is worth discussing briefly the need for this type of professional telecommunications engineer whom we are consider- ing. Following is a listing of significant elements of the telecommunications industry that already need - or will in the near future - need additional professional telecommuni- cations engineering talent. - Common Carrier: Domestic common carriers like ATT, independent telephone companies, specialized common carriers for data and teleprocessing like Datran, MCI, Western Union, etc. The case of ATT is unique and will be discussed separately below. International common carriers like COMSAT, ITT, RCA, Western Union International, etc. - Federal Government Departments and Agencies: Defense Communications SystemsTincluding the Army, Navy, Air Force, DCA and a number of special agencies. Federal government departments other than Defense such as Commerce, HEW, DOT, Justice, State, HUD, NASA, etc. - State and Local Governments: All of the states, and some with special needs, such as Alaska and Hawaii. There

75 will also be special needs for professional analysis and system engineering in many large municipalities with their unique telecommunications problems. - Manufacturing Industry: This is the area of the greatest need for telecommunications professionals. The firms with such needs include giants such as Western Electric, RCA, ITT, GE, GTE, IBM, Texas Instruments, Raytheon, Motorola and others, as well as the thousands of smaller firms which make up this dynamic industry. - Broadcasting: This field includes the television net- works, the FM and AM radio domestic and international broad- casters, all of whom will continue to need professional telecommunications engineering talent into the near future. - CATV: The related field of cable television is emerging with a projected growth of l0-25% a year over the next l0 years depending on the regulatory climate. This will call for a large amount of professional engineering talent, if it is to fulfill even a portion of its promise for the industry and society. In summary, it seems evident that the need and demand for high quality professional engineering talent in telecom- munications is, and will remain, very high, as we face our own anticipated domestic growth and the more sophisticated technological challenges from abroad. VI. Special Case of ATT ATT represents a very large segment of the telecommuni- cations industry from almost any standpoint - be it planning, system architecture, operations, or research and manufacturing. Its need for properly trained professional personnel is very substantial and vital. But ATT depends on the schools and universities only in part, namely to provide personnel with the right intellectual potential and a good basic scientific education. The specialized training necessary to match ATT personnel to their careers in ATT is provided in the organization through internal academically oriented training courses. Also, in cooperation with schools and universities, ATT encourages and subsidizes MS, PHD and post doctoral studies. This method insures a good supply of competent and well trained professional personnel. There is another characteristic of ATT professional per- sonnel which has a profound impact on the whole industry. To a great extent the electronics industry has depended on

76 the mobility of the engineers to spread new technology and techniques. This has been particularly true in the aerospace portion of the industry where it is unusual to find an experi- enced engineer who has not worked for more than a half dozen organizations. This has been the method of professional advancement - going from job to job and moving upward on the responsibility and salary scale. As a rule ATT telecommunications engineers and scientists do not move out of ATT very much. Strictly pro- fessional turnover in the Bell System is very substantially lower than in any other reasonably comparable telecommuni- cations-electronic organization. The facilities at ATT are unequaled and opportunities for professional advancement are very good. This means that the rest of the telecommunications industry must take care of its own professional training as best it can, depending on internal resources, on graduate professional training in the schools and universities, and on practical on-the-job training or no training at all. There is one other characteristic of the ATT training in the industry which is very interesting. As a rule elderly electronic engineers are not in demand in industry because, except in management areas, they are considered not up-to-date. In the case of some ATT retirees however, their technical know-how seems to be in considerable demand, because they have specialized knowledge in telephone networks, switching, traffic studies and related areas where knowledgeable pro- fessionals (outside ATT) are very scarce. A number of them have established themselves as consultants in telecommunications to the industry's benefit, as well as to their own. This is only happening because there are not sufficient numbers of professionals being trained in this area outside of ATT. The more that competition is encouraged in the industry, in the data communications, CATV, specialized carrier, state and local, and international trade areas, the more severe will this problem become and the greater the need for well-trained telecommunications professionals. VII. Some Tentative Conclusions The training and education of telecommunications engineers needs to be reexamined, revised and broadened to take into consideration: - the huge growth of the industry; - its increasing complexity and sophistication; - its economic and social importance to the nation.

77 This training will emphasize, in addition to science and technology, system architecture, system engineering, applications engineering, and the social and economic impact of all aspects of telecommunications. The new curriculum, including some of the material in Section IV above, would in all likelihood require an additional year or more of professional training. A more detailed study, analysis and further consultation with concerned professionals will be required before a significant recommendation can be made along this line. Many schools are becoming aware of this problem and are beginning to revise their electrical engineering courses to give a more prominent place to telecommunications. The National Science Foundation (NSF) should help and support this tendency. One form of such help might be for the NSF to take the lead in establishing a well qualified task force with repre- sentation from the universities, industry, professional societies and the major users of telecommunications, including the Depart- ment of Defense. This task force would study in depth the problem of the training and education of telecommunication engineers to meet our national needs. The task force would end up with a set, or possibly several alternative sets, of specific recommendations for all professional schools and universities for future action on a voluntary basis. An additional part of the task force work would be to study in some detail the scope of training of telecommunications professionals in other technically advanced countries, such as England, Germany, France, Japan, the USSR, etc., in order to determine whether we can learn something from abroad. Though in most cases the foreign telecommunications industries are very substantially differently organized from ours, they are beginning to face many of the same problems as we are. VIII Graduate Centers for Telecommunications Policy Studies All of the above pertains to professional engineering training, including consideration of industrial and social problems as they impact the engineering field. It is not meant to concentrate on the primarily social, societal, economic and legal issues which are important to society. For the latter, the National Science Foundation effort to establish Centers for Telecommunications Policy Studies in several universities and not-for-profit foundations looks

78 like an important step in a desirable direction. Such centers,when effectively implemented, could do justice to the many policy issues and implications which involve society, the general public and the nation in a very significant way. Armig G. Kandoian Telecommunications Consultant Ridgewood, New Jersey

Next: TERMINALS, SENSORS AND DISPLAYS »
Telecommunications Research in the United States and Selected Foreign Countries: a Preliminary Survey. Report to the National Science Foundation Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF
  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!