This html article is produced from an uncorrected text file through optical character recognition. Prior to 1940 articles all text has been corrected, but from 1940 to the present most still remain uncorrected. Artifacts of the scans are misspellings, out-of-context footnotes and sidebars, and other inconsistencies. Adjacent to each text file is a PDF of the article, which accurately and fully conveys the content as it appeared in the issue. The uncorrected text files have been included to enhance the searchability of our content, on our site and in search engines, for our membership, the research community and media organizations. We are working now to provide clean text files for the entire collection.
Introduction: Although it is not a phenomenon that is particularly noticeable to us at any given moment, we are living in an era of unprecedented technical change. Rapidly evolving technology affects our lives in countless ways, but we accept it and adapt without particular trauma or even awareness of the significance of this evolutionary process. The pocket calculator, color television, plastic products, and space exploits seem quite ordinary in private life. In the naval profession, we are perplexed by problems of obsolescence brought about by rapid technological change, but we don’t have a sense of blurring progress there either. This is consistent with the amazing human capability to adapt. However, our characteristic acceptance of change can allow unwarranted complacency about long-term effects. In the military world, this can be a serious oversight.
There are some changing things, however, that promise noticeable shock value because of their extreme rate of variation. The energy crunch, as it closes in, should cause very perceptible problems with sharp price rises, vexing fuel shortages, and regular public crises of various sorts to attract our attention. Shortages of raw materials and food may also accelerate in a manner which will affect our lives quite obviously, both nationally and worldwide. Another kind of change which may have an avalanche effect on our professional and private lives is the digital computer explosion, triggered by the development of the microprocessor and its associated microcomputer. Much has been written about the microprocessor in newspapers and magazines, but its impact on military operations has not been the subject of much speculation. Such speculation is the subject of this article. Digital technology is moving with incredible speed, especially in the production of smaller but still powerful devices. A “generation” in the field of digital hardware is only about 15 months, bringing about significant changes in capability and cost. The potential applications are virtually unlimited, and they are certain to have profound impact on naval warfare.
What is a Microprocessor? Physically, it is an object smaller than a dime. In its package, it is about the size of a stick of chewing gum. The base structure is a thin wafer of silicon on which is “printed” a microscopically fine array of conducting strips, in many layers, together with precisely placed specks of semiconductor materials. It costs millions of dollars to accomplish design and production engineering for one of these chips, but once the machinery starts rolling, they can be delivered to the consumer for $ 10 to $20 each.
Some of these chips, which actually include tens of thousands of interconnected electronic devices, are complete and capable digital computers. They can store and execute programs in a manner analogous to the refrigerator-sized computers with which most of us are familiar. Their capabilities, in terms of memory, instruction sets, and speeds, are increasing by leaps and bounds, and the cost per unit is going down at the same time. The economic message here is that efficiency of computation in terms of hardware cost is becoming less and less important. We can now consider using sophisticated digital processing in functions for which it would have been considered ridiculously extravagant 10 or 20 years ago. Most middle-grade naval officers can recall when “core capacity” was a limiting feature of systems using digital computers. There were desirable combat system functions which could not be performed in given ships for lack of core space, and there were constant efforts to reinvent or restructure computer programs
Tiny chips, no less than huge ships, will be needed if the Navy is to retain its technical leadership in naval warfare. Two of the tiniest, seen flanking a boatswain’s pipe for actual-size comparison, are the computer-on-a-chip (Intel 8748), left, and the interpreter-on-a-chip (National Semiconductor). All three are superimposed over a micro-photograph of the circuitry of an Intel 8080 microprocessor.
to make more efficient use of the available capacity. With these circumstances, one could consider use of digital computers for only the most elegant and important problems. Something as mundane as monitoring an air conditioning system was unthinkable.
The standard Navy tactical data system (NTDS) computer of the 1960s had core capacity of 32,000 words. This was respectable but not enough to satisfy the needs of even that decade. System designers strived for more computer capacity through additional equipment, and competing requirements sometimes “bumped” desirable capabilities out of shipboard computer programs. Now, a complete computer is available on a single “chip” with 1,000 words of memory. With another tiny chip, 4,000 or 8,000 words of additional memory can be had for less than $50. By 1980, we can expect to have microcomputers operating at 100 megahertz rates, with 16-bit word length, with each chip having many thousands of words of core—approaching the capacity of the AN/USQ-20 now in the fleet. The core capacity problem will fade rapidly. We will turn from concern for how to fit things in to wondering what else should be done with computers.
The basic thing that a digital computer can do is arithmetic. It can add, subtract, multiply, and divide to solve mathematical problems as do the hand calculators with which we are now all familiar. Although it may not be obvious, this almost limitless arithmetic capability also allows the machine to make complex logical decisions, to analyze acoustic and video waveforms, to draw graphic diagrams, to translate one computer language to another, and to accomplish many other functions that would be difficult or laborious to a human operator. Step by step, the actions are elementary, but the strength of the digital computer is that it can do millions of simple steps each second, and these millions of simple actions can be designed to accomplish very impressive tasks. In addition, it can receive information in its arithmetic processor for comparison with data stored in memory. It can also be ordered to search accessible memory for certain kinds of information and to inform the requestor as to what is available and perhaps what it means. These faculties can be applied to perform control functions if desired.
An analog example which might be appreciated is the standard home thermostat which has a “program” of sorts stored mechanically. The user “inputs” the desired temperature and, with a switch, whether he wants to heat or cool. The thermostat measures the existing temperature and makes a logical decision whether or not to turn on the selected machine. Once the machine is functioning, the thermostat monitors the situation until the “input” value is achieved, and then shuts off the machine. The thermostat can be made a bit more sophisticated by building in some tolerance limits, control for both heating and cooling, and some of the more expensive models include clocks which modify operation as a function of time. That’s about as far as one would think to go with a mechanical thermostat, but if the job were being done with a digital processor, a designer could readily instrument it to take into account humidity and add other useful refinements to comfort control. Digital logic would enable the designer to compare almost any number of conditions and provide for complex logical decisions. Environmental control could be made far more sophisticated than could be reasonably accomplished with an analog device.
Another analogy might be instructing a furnace operator or a gate guard on how to do his job, based on some finite number of circumstances or requirements. As long as his instructions are sufficiently comprehensive and understandable, the worker is expected to perform accordingly. Within limits, you can do the same thing with a microcomputer and a set of sensors. The difference with the digital device is that it will not get tired, distracted, or discontented. Neither will it apply judgment beyond its given instructions, which may be an advantage or a handicap depending on the situation.
Applications: Application is the crux of the digital evolution insofar as the Navy should be concerned. Imaginative people are predicting extensive future applications in all areas of human endeavor, and it is certain that military applications will be far- reaching. At the moment, however, we have just begun to scratch the surface. We are moving slowly, perhaps too slowly, for several reasons. Some of our problems are as follows:
► Immediate experience with digital systems is not all good. Many development programs have had discouraging problems related to software development. Neither has computer hardware been completely reliable, and there have been memorable casualties associated with accidental loss of stored memory or programs that “bomb.” With these experiences vivid in the recollections of contemporary naval leaders, there is understandable hesitance to leap forward into greater dependence on digital systems.
► The software costs and core limitation experiences of early digital systems have conditioned many of us to react with caution toward digital-dependent weapon systems. It is a fact, however, that the mem-
n
ory core limitations of earlier technology contributed directly to associated software complications and expense. Programmers were driven to extremes of cleverness and complexity to write their programs in the absolute minimum number of instructions. Simpler approaches, using core more liberally, require less human creativity, are more readily managed through development, take less time to complete, and are less likely to have serious operating bugs.
► A major drawback in advancing digits in the Navy is poor perception of the applications of better and cheaper digital processors. Many of us tend to think on the basis of our experience wherein computers were large and complicated equipments, housed in special compartments and used only for the most important combat functions. It is difficult to imagine a future naval ship having hundreds of tiny computers, distributed throughout many different spaces, and given no more maintenance attention than a fuse box or a motor controller. Because it is hard to think of computers as ordinary devices, the potential user has an understandably dim perception of new functions they might perform in the operation of a warship.
► Digital technology is moving rather fast for the military technical establishment. The material acquisition organization is geared to development cycles of five to seven years from concept formulation to the beginning of production. The microprocessor industry, on the other hand, can accomplish a major development effort over a period of only months from inception to putting the device on the market. Our technical naval officers and civil service engineers are hard pressed to keep abreast of this technology.
► It is a fact that the Department of Defense does not represent an attractive dollar market for microprocessor production. Some of our larger development programs might use only a few dozen microcomputers per system. At $10 a chip, this does not represent a market worthy of the time of a sales engineer. He goes instead, to the toy, auto, and appliance industries where sales are already in the millions of units. The significance of this situation is that the Navy research and development establishment cannot wait for industry to sell it the product, as it has been accustomed to waiting with regard to new developments for many years.
► The DoD acquisition process is slow to adapt to new technical situations. It took decades for procurement rules to cope with the proliferation of different kinds of second-generation digital computers, and current mechanisms are geared to large and costly machines. The same procurement and control mechanisms, unfortunately, are generally applied to
the progressively smaller and less expensive computers. It may take a number of years before $10 computers can be purchased without months of administrative processing.
► A basic problem is the relatively small number of people in the Navy who really understand computers and digital processing. There is reluctance to rush into the purchase of digitally controlled systems, however attractive, when management has little confidence that the operators and maintenance people in the fleet will be able to cope with them.
Assuming that our institutional barriers to progress can be overcome, let us consider the broad potential of these new families of digital devices. Some useful attributes may be grouped as follows.
To Manage Information: The digital computer or processor has the capability to store and retrieve information. The fact that it does it in binary digits is immaterial. Almost any kind of information that can be written or illustrated can be stored in digital
SIO.OOO to
SIOOJOOO
MICRO COMPUTERS
(LSI)
1976 Large scale
integration 10,000 transistors on a chip
(VLSI)
Very large scale integration up to 100,000 transistors on a chip
PERSONAL COMPUTER 1980
form. Having stored some information, the device has the capability to sort it according to rules which it may be given, or to search the collection to find, process, or return information that is desired. Essential to these functions, of course, is a method to “output” information in some useful manner. So the digital processor, through ancillary equipment, is usually rigged to tell, send, print, or store selected information.
As an example of information management that might appeal to people who have spent a good portion of their lives standing bridge watches, imagine a microprocessor-controlled quartermaster’s log. Instead of writing illegibly with a stubby pencil, sometimes in the dark and often late, the quartermaster of the watch might have a small console with a limited number of labeled (and lighted) information entry buttons and a typewriter-like keyboard. A selection of standard entries to the quartermaster’s log could be preplanned with buttons labeled “course change," “speed change,” “casualty,” “emergency drill,” “change of watch,” “navigation position,” “wind direction/speed,” “air temperature,” and so on. Although there may seem to be a large number of possible entries, 20 or 30 labeled buttons could handle the great majority of log entries. In addition to his labeled entry buttons and keyboard, the quartermaster would have a small visual display to indicate the content of the material which is about to be entered into the record.
Upon the ordering of a course change, he would not need to enter the time; that would be done automatically. He would touch the entry button labeled “Course change,” punch the three numbers indicating the new true course, observe the validity of the entry on the display, and then touch the “enter” or “log” button to have that information permanently recorded on a magnetic storage device. With a change of watch he would press the “change of watch” button, type in “section 3” and perhaps type the name of the officer of the deck. There could be labeled buttons for “captain on the bridge,” “captain off the bridge,” “navigator on bridge,” “navigator off bridge,” etc. All these details of bridge activity would be logged in a legible and accurate chronological record with minimum of time, training, and effort. The quartermaster would have to type in proper names, and he would occasionally face the need to describe a situation not covered by his entry buttons. For this he would use the typewriter keyboard. For writing up the official deck log, required types of information could be filtered by preprogrammed selection of quartermaster of the watch entries. Other data series, such as successive courses and speeds, could be similarly sorted out and even plotted from information entered into this automated quartermaster’s log.
Another application in the field of information management which might appeal to an executive officer would be help in the preparation of watch, quarter, and station bills. The inputs to this process are the stations to be manned under various conditions, together with the names and qualifications of the people available. A digital machine can do this kind of tedious sorting and matching quickly and accurately. Problems resulting from unexpected receipt, transfer, or unavailability of a key man, could be instantly researched and the machine could make recommendations for adjustment.
With qualifications of people on file in such a data bank, watch officers would be able to review the qualifications of the people with whom they are teamed on the bridge or at battle stations. The training officer could have accurate forecasts of upcoming requirements for school quotas or other forms of special training. The process of making out watch bills could also be greatly aided by a small computer programmed to shuffle names, requirements, leave schedules, fair shares of holiday and weekend assignments, and the many details that must be taken into account in the watch bill writing process. This is possible today with existing computer installations, but the microcomputer will make it economically practical.
Monitor Conditions: The digital computer has a noble capacity to “stand watch.” Given a set of conditions which should obtain, the digital machine can reliably and tirelessly monitor those specified conditions and compare them with given standards. The comparison logic may be quite complicated, with any number of ifs, ors, ands, or buts. When the situation varies from the norm, the processor can sound an alarm or order correction of conditions. A comparable monitor function is often assigned to a man, i.e., a watch standee;-To the extent that human judgment is neither required nor desired in selected situations, there is real potential for replacing people with digital monitors. The bonuses might include more reliable performance and morale improvement by excusing intelligent people from boring assignments. The hardware reliability would probably exceed the overall performance of an assortment of human beings with their varied personalities, experiences, interests, training levels, attention spans, intelligence quotients, personal distractions, fatigue, and the like.
As an example in this area, consider a watchstanding problem that must be as old as any navy, the “anchor watch.” This is an individual charged with observing the ground tackle to assure, as best he can, that the anchor is holding and that the ship is secure. Unlikely as it is that an anchor will fail to perform, the chance is still too great to ignore, especially in a large ship. There can be no excuse for going aground because no one noticed that the anchor chain had parted or that the anchor was dragging. On the other hand, simply having a man forward does not necessarily assure safety. The average sailor is not equipped to determine how much strain is on the chain or how much strain there should be, and only a veteran seaman is competent to judge when an anchor might be dragging.
A small microcomputer could help in this area if coupled to appropriately placed strain gauges in the deck structure near the chain stoppers. Strain gauges might be arranged to read the actual tension on the anchor chain and possibly to compute the direction of tend. With programmed thresholds for maximum and minimum strain and the proper directions of tend, a monitor could allow high confidence at a remote location that the anchor was doing its job. This is a fairly trivial idea, but it could save manpower, and it would probably do the job better than the average assignable individual.
To Compute Mathematical Solutions: Computers do “compute” even though they are often designed or used for nonmathematical purposes. They are ideally suited to the mathematics of weapons control, performing calculations with speed and accuracy beyond human comprehension. Digital processors can compute courses, speeds, CPAs (closest points of approach), times of flight, times of intercept, lead angles, turn rates, and all the other calculations necessary to engagement of targets. In addition, they can generate commands which are transmittable to missiles equipped with compatible interpretive devices. We have already advanced beyond sending “steering orders” in analog fashion to missiles. We now send “where to go orders” in digital format. The missiles
The “computer on a board” is an impressive achievement and is appearing now in military equipment. The “computer on a chip" is already available commercially, and a “computer system on a chip" is just around the corner.
****'♦»"
COMPUTER ON A BOARD
1976
7,000 TRANSISTORS
PER CHIP
COMPUTER ON A CHIP
_ 1977
30,000 TRANSISTORS
COMPUTER SYSTEM ON A CHIP
ANALOG
CPU | MEMORY | 1980 | ||
A/D | D/A | DIGITAL I/O | 100,000 TRANSISTORS | |
J ------ ' | u | DIGITAL | ||
decode digital commands and maneuver accordingly to positions in space. Microprocessors are advancing such capabilities.
Another attractive possibility of the microcomputer might be termed “local fire control computation.” Gun mounts and missile launchers might be given no more than continuing line-of-sight data and range to a target. A tiny microcomputer, located right on a gun mount, could compute the lead angle for its particular gun ballistics and generate local orders for pointing and firing. This could have the attractive benefit of dispersing fire control computation capability, reducing the amount of wiring and interconnections required, and greatly simplifying “fire control plot.”
Aside from solving fire control problems, digital processors can also be used to compute optimal courses of action as an aid to human decisionmaking. In cut-and-try fashion, but with thousands or millions of iterations, a digital processor can either recommend or order actions which require the least time, maximize probabilities of mission success, make most efficient use of available resources, increase the probability of survival in the face of certain types of enemy weapons, and solve many other kinds of tactical problems. We may know from experience that there is likely to be a “best way” to respond to a given situation, but we do not ordinarily have the time to explore many alternatives and assure ourselves that a proposed tactical decision is optimum. Digital computers are already of great help in these kinds of decision-making processes, and the microprocessor will extend this kind of help to areas where it is not now practical. Through this advancement of greater computational capability at lower cost, it is practical for a commander afloat to have his own “war game machine” whereby he interactively refines his plans for deployment and use of forces. He can do this off-line, on a small machine, without affecting operations in progress.
Other computational possibilities for improving shipboard operational performance might be in peaking power plant economy, using as data inputs the wind, sea state, humidity, air temperature, sea temperature, and desired speed. A microcomputer could yield optimal recommendations for engine rpms, propellor pitch, fore and aft trim, allowable operating temperatures, and so on. These kinds of refinements are ordinarily done with handbooks and curves supplied to the operating engineers, but their inputs are limited and coarse approximations are necessary. A microcomputer could be “on line,” constantly recomputing with changing inputs and recommending improvements to the controlling personnel.
A vital area where rapid computation of relatively complex problems can sometimes be important is damage control central. Given a serious hull casualty, a dedicated computer could analyze quickly the effects of damage, reporting in terms of affected piping and electrical lines, changes in stability, probable loss of fuel, and the like. A computer could search files of hull systems information associated with a given area of the ship much faster than any human being. It could be programmed to recommend valves to be shut and alternate power sources to be picked up. It would do the same things that men attempt to do now with the damage control diagrams, but the answers would be almost instantaneous, subject only to verification by human operators.
Significance of the Digital Navy: The “bottom line” is fleet capability. In achieving and maintaining combat superiority over potential adversaries, technology must be the trump suit of the U. S. Navy. We cannot hope to outman our potential enemies or even to have more ships, planes, or missiles. We can, however, maintain a combat advantage by means of superior performance in our systems. Our weapons may be fewer and no larger than those of our enemies, but they have to be better in every other respect. We are well into the age of the “smart weapon,” and the digital processor is unquestionably critical to system smartness. This technology, however, can yield benefits in many other shipboard functions apart from the combat systems. Digital processors will eventually find their way into virtually every machine and piece of equipment afloat.
Aside from specific examples of possible new microcomputer applications in shipboard operations and systems, there remain to be explored unimaginable human initiatives with these new machines. As the steam engine started the industrial revolution by multiplying the muscle of man, the digital processor promises to give new leverage to the mind of man. We are in the early phases of a technical revolution comparable to the impact of the transistor. Personal computers will proliferate in America like televisions and radios. Hundreds of retail shops have already sprung up across the country to sell small computer systems for private use. Sales of personal computers were expected to reach $50 million in 1977 and a billion dollars by 1985. It is estimated that by the year 1980 there will be programmable computers in about 10 million American homes, and in another ten years almost all American homes will have them. Thousands of people will inevitably invent new solutions to their personal and professional problems
based on these devices, and this process should be encouraged wherever it occurs.
The significance of this personal computing trend is that there is a ground swell of public interest to learn and use this technology. It will become part of the American socioeconomic fabric and consequently spread to all other countries of the world. We will soon be afloat in small computers, like them or not. It seems most important that the Navy take the initiative to maintain technical leadership in their application to naval warfare. There is, as noted earlier, too little industrial incentive to invent Navy uses for microprocessors because the DoD dollar market is small compared with the civilian market. The imagination will have to come from within the services. We should educate Navy people in the potential of microcomputers. We must search out our unique applications and engineer them. This will happen, sooner rather than later, if we somehow combine knowledge of the Navy’s problems with knowledge of the microprocessor. This suggests a deliberate educational program for Navy personnel ashore and afloat and a vigorously prosecuted Navy research and development effort. The magic will not happen quickly if the operating people wait for the shore establishment to do all the thinking. For all the brave labels we have on our laboratories, and despite a long list of dedicated Navy-associated contractors, people who do not go to sea cannot appreciate fine grain operation and maintenance problems. They will not automatically identify microcomputer applications for warships with acceptable coverage. They need to hear ideas from the operators. Once the ideas are conveyed, the research and development establishment can do wonders, and that is as it should be.
However, ideas for research and development should be based on technically informed judgments as well as perceived needs. A sailor will not generate a bright idea for application of a microcomputer if he does not know what such devices can do. The Navy could and should take overt steps to inform the seagoing population of this new technology by literature, films, and hardware. Every ship could have several “personal computers” for use by crew members, allowing opportunity for those who lean toward these kinds of gadgets to exercise their imagination and perhaps generate useful application ideas at the operator level. These machines, complete with displays and mass memory facilities, can be had for the price of a color TV. Videotape courses are available, computer hobbyist magazines are on the newsstands, and other instructional literature is abundant.
This critically important technology is racing ahead; the hardware and software are in hand; the
Navy can benefit greatly by intelligent exploitation of the product; and we have a moral obligation to do so. We can sit back and let the goodness seep in with 10 or 20 years phase lag, or we can reach out boldly to take advantage of it now. The approaches should be both formal and informal. Our research and development offices, laboratories, and contractors should be vigorously engaged in learning how to use microprocessors, and our fleet people should be encouraged to get involved at all levels by making the tools available to them through structured educational programs or simply making inexpensive computer systems available for personal use. The cost is small—the benefits incalculable—and the penalty for lethargy severe.
0 Commander Keen was commissioned at the NROTC Unit, University of Virginia, in 1946 with a degree in naval engineering. As an unrestricted line officer, he served in amphibious and mine force units during the early years of his career, commanding the USS Dukes County (LST-735) and the USS Esteem (MSO-438). He completed the engineering electronics curriculum at the Naval Postgraduate School at Monterey, receiving an M.S. degree in 1955. His last sea tour was as missile officer in the USS Canberra (CAG-2). This was followed by three years as Special Technical Assistant to the Director of the Surface Missile Systems Project. Since retirement from active duty in 1965, Commander Keen has been an assistant to the Director of the Applied Physics Laboratory, The Johns Hopkins University, and involved in a wide variety of technical projects in support of Navy research and development programs. His article "A New Kind of Navy” appeared in the January 1978 Proceedings.
Mr. Perrine is an assistant to the Director, Applied Physics Laboratory, The Johns Hopkins University. He has enjoyed a very productive career in radio communications, electronics engineering, guided missile development, and related technical fields. A graduate of the California Institute of Technology in Experimental Physics and elected to Tau Beta Pi, Mr. Perrine worked for the Hughes Aircraft Company, Fairchild, the Bendix Corporation, and Convair-Gencral Dynamics over a period of 35 years. He was Executive Vice President of the Pomona Division, General Dynamics Corporation, where he led the development and production of Terrier and Tartar missiles for the Navy, prior to joining the APL staff in 1973 In recent years, Mr. Perrine has been increasingly involved in examining the potential of microprocessors in the solution of Navy operational problems.
HMr. Hazan obtained his bachelor of science in electrical engineering from the Royal College of Science and Technology in Britain. He did his graduate work in computer science at the University of Maryland. Mr. Hazan is on the Director’s Senior Staff at The Johns Hopkins University Applied Physics Laboratory, where he leads the laboratory’s microprocessor program. He is on the board of governors and is Director of the IEEE Computer Society for Micro and Minicomputers. Prior to joining APL, Hazan was Technical Director of the Singer Company in Maryland, responsible for the development of large-scale Navy simulators and training systems. He is the author of numerous papers on computing, and has chaired several national and international computer conferences.