About Ralph Benjamin
Ralph Benjamin was born on 17 November 1922 in Darmstadt, Germany. Due to the pre-war hatred toward the Jews, his parents sent him to a boarding school in Switzerland in 1937. Since the threat of Germany escalated, he had to stay at a refugee camp in Switzerland and was sent to England in 1939 by the English refugee organization. While attending Ellesmere College in Shropshire, he decided to switch to Science and successfully fulfilled his academic requirements. After his plan to attend an American university was thwarted by the war, he got a job as an electrician's mate and earned a scholarship to Imperial College. From 1944 to 1946, upon completing his education, Benjamin worked for the Scientific Civil Service at the Admiralty Surface Weapons Establishment (ASWE). He moved to the Admiralty Underwater Weapons Establishment (AUWE) and was involved in developing information, intelligence, and military (torpedoes, submarines, etc.) technologies. After the seven-year's work at the AUWE, Benjamin started a new career at the Government Communications Headquarters (GCHQ) at Cheltenham, UK. He carried out Signals Intelligence tasks and developed a speech security system and an encryption system. When he retired from the GCHQ post, Benjamin joined NATO and got an appointment to the Supreme Headquarters Allied Powers of Europe (SHAPE) Technical Center as Head of the Communications Technique branch. Upon completing his term at NATO, Benjamin was involved in academic activities at various universities in the UK. He also held several memberships at advisory councils and boards, including the Defense Scientific Advisory Council. Ralph Benjamin received a number of academic and other awards for his decades-long contributions.
In the interview, Ralph Benjamin not only provides his personal contributions and commitments to various institutions but also offers broader reflections on intellectual activities and management issues at these organizations. In addition, he shares general thoughts on the institutions that he worked for and his efforts to challenge the problems and to improve productivities of the staff involved. Although Benjamin's career was devoted to the Civil Service, he also performed academic activities at several universities, participating in research projects, sharing ideas, and helping young students achieve their goals. He shows in the interview how much he values the importance of personal contributions and direct interactions with people at the working level. Benjamin shares his experiences of managing individual and tem projects/ programs as well as of running an R&D establishment. The interview concludes with Benjamin’s final remarks on his work ethics.
About the Interview
RALPH BENJAMIN: An Interview Conducted by Peter C. J. Hill, IEEE History Center, 16 September 2005
Interview #465 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc., and
Rutgers, The State University of New Jersey
This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.
Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, Rutgers - the State University, 39 Union Street, New Brunswick, NJ 08901-8538 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.
It is recommended that this oral history be cited as follows:
Ralph Benjamin, an oral history conducted in 2005 by Peter Hill, IEEE History Center, Rutgers University, New Brunswick, NJ, USA.
Interview: Ralph Benjamin
Interviewer: Peter C. J. Hill
Date: 7 July through 16 September 2005
Place: Bristol, England
This is an IEEE UKRI Section oral history recording made by Peter Hill, Communications Chapter, London, England. Peter Hill is currently based at Cranfield University, Defense College of Management & Technology in the Defense Academy of the United Kingdom, Shrivenham. The recordings began on July 7, 2005 and were made in Bristol, England. The interviewee is Professor Ralph Benjamin, CB, Ph.D., DSc, DEng, FIEE, FCGI, FREng. He entered the government scientific service in 1944 and worked in the areas of Naval Command and Control, sonar and communications, finally rising to the position of Chief Scientist of the Government Communications Headquarters (GCHQ), Cheltenham. He went on to work at Supreme Headquarters, Allied Powers in Europe and SHAPE (in NATO), retiring in 1987. He is now a visiting academic professor and consultant. Professor Ralph Benjamin, thank you very much indeed for volunteering to give us an IEEE oral history. May we please start with something about your early history, including how you moved to England? Then later we can discuss your early career at the Admiralty Surface Weapons Establishment (ASWE) in England.
I was born on November 17, 1922 in Darmstadt, Germany. There is nothing worth saying about my elementary school. I went to a secondary school, which the Germans call a gymnasium, where I did pretty well in all the subjects. In the “classics”, I enjoyed reading Caesar's Gallic Wars and Virgil's Iliad in Latin, and enjoyed even more greatly The Odyssey and especially the adventure epic of Xenophon’s Anabasis in Greek. In the playground I enjoyed the general rough and tumble, but being Jewish I increasingly had trouble with some of the children from other classes – never from my own – particularly on the way to and from school. In 1937, when my parents decided that this was really getting too much, they sent me to a boarding school in Switzerland. That boarding school had an English department and during the preceding Summer holidays I had a few lessons with an elderly lady who had once lived in England. In Switzerland I studied for what nowadays is called General Certificate of Education O levels, but in those days was called Oxford School Certificate. All of the subjects were taught by an adventurer from New Zealand. He was assisted by a young man who had been expelled from Oxford University. They both were very good. Unfortunately, after a few months the Germans cut off the payment of school fees. Therefore, I had to leave the school and go to a refugee camp. I took my schoolbooks with me and at the end of the year the school allowed me to sit for the exam, which was invigilated by the British Consul and in due course marked at Oxford. I got either credit or distinction in all seven subjects I took, including English Language and English Literature.
I had to stay in the refugee camp until March, 1939, when the English refugee organization, at very short notice, got me a stand-by a place in a plane going from Switzerland to Croydon. I had a widowed aunt living in England, but she did not know I was coming. Therefore, when I arrived at Croydon, South of London, with my two suitcases I had to make my way from there across the public transport system to Edgware in North London and then by foot to her home. I stayed there until the summer of the following year when I managed to get a scholarship to Ellesmere College in Shropshire. I had a very difficult interview with the headmaster, who was impressed that I was well ahead of anyone there in Latin and Greek, and was keen for me to follow that. However, I had got a bee in my bonnet that I wanted to switch to Science. He thought I was completely crazy, because I had done no Chemistry, had a big gap in Mathematics – having learned no Calculus or Coordinate Geometry – and had big gaps in Physics – having taken no Optics and so on. Eventually I got my own way, which meant that I had to attend a few lessons in classes from the Third Form up to the Sixth Form. However, most of the time I had to study on my own. In practice, I benefited from this, because being on my own I managed to be ready for the exam after one year, rather than after the usual two years in the Sixth Form, for what was then the Higher School Certificate and nowadays is termed A Levels.
In the meantime, on the basis of my ordinary School Certificate results, I obtained a scholarship to an American university. However, by the time I got an American Visa, the war had broken out, and so I could not get a place on any ship. Therefore, I stayed and got myself a job as an electrician's mate. In the beginning I was employed to install lighting and heating in air raid shelters in North London. Then I managed to switch to Boosey & Hawkes, a company that had been making electronic organs. I helped them to switch over to the manufacture of Blenheim bombers. Eventually I ran into the father of one of my school friends from Ellesmere. He found out what I was doing, and through his initiative I got a private funded scholarship to Imperial College. I became enrolled in a wartime scheme called the Hankey Scheme for accelerated degrees in technical subjects needed in war industry and the Forces. During my first year at Imperial College, I also played chess and boxed for the College: two of the few extracurricular activities that were still going on. However, due to the bombing in London, they too stopped after a year. That may have been a good thing for my academic work.
Throughout my studies, I found it much more satisfying to work out results from first principles, rather than to memorize end results. Indeed, even in school I had a reputation for working out my own solutions to problems. School teachers and lecturers found it a bit irritating, because if I got the right results, as I normally did, more quickly but by a method that was not self-evident to them, they found it much harder to mark the work. In one of the lab experiments I developed what is nowadays known in radio communications as a single-sideband (SSB) mixer. I did not realize, and neither did my lecturers, that it was going to be several more years before this was re-invented in industry. Our finals took place at the height of the V1 flying bomb attacks. The exam was in a glass-roofed drawing office on the top of the City and Guilds building. The whole exam took place during an active air-raid warning. When air raid warden, posted for this purpose on the roof above us, heard a flying bomb start its terminal dive, he blew a whistle, and we all had to dive under the tables. The invigilator had a stopwatch, so that we got extra time at the end to make up for that. I finished with First Class Honors, and Imperial College wanted to keep me in a lecturing appointment. However, what one did at this time was the decision of the Ministry of Labor and National Service, and they wanted me to go to Aberdeen University to teach RAF (Royal Air Force) technical cadets.
I wanted to join the Navy, and, I was eventually assigned to the Royal Naval Scientific Service, to work on the very new subject of Radar. So, in 1944, I was assigned to the Admiralty Signals Establishment (ASE), and eventually stayed in it until 1964. During this time it changed its name repeatedly. Radar was really its dominant activity. They were going to call it the Admiralty Radar & Signals Establishment, and got as far as printing the first letterheads with ARSE, before deciding it might be more tactful to make it the Admiralty Signals & Radar Establishment (ASRE). Later still it became the Admiralty Surface Weapons Establishment, in a merger with another Admiralty Department. Radar, being a new field, had predominantly new people in it. Many of them were of course from British academia and industry. However, we had an interesting mix including some free French, Norwegian scientists, one or two South Africans, one or two Americans, and quite a lot of very able Poles, and we were supported by a number of NCOs (Non-Commissioned Officers) from the Women’s Royal Naval Service, to act as very popular and attractive lab assistants. There was of course no relevant literature on radar. One of the most useful reference books on radio was the Admiralty Handbook on Wireless Telegraphy written by a certain Louis Mountbatten when he was a Lieutenant Commander.
We were very fortunate to mix fairly intimately with so-called Naval Applications Officers, primarily operational ones, to make us understand the navy’s operational problems, but also technical ones to make us understand the problems of operating and maintaining equipment at sea. I started off as a temporary Experimental Assistant, Grade III. Because I had a 1st, I got a bit of extra pay, so I finished up with £4 a week. Eventually I became a temporary Experimental Officer and climbed up all the way to £6 a week. We were all working very hard, and I got very involved in my work. Therefore, as often as not, I forgot to collect my pay, but they kept it for me until later weeks. There was not a lot of spare time, but I did help to run a local scout group, the First Milford. On one occasion, cycling to the scout group, I got run into by a military vehicle of the Canadian Army, which was driving fast on the wrong side of the road – the right side by Canadian standards – and I finished up in hospital with concussion and a broken jaw. However, as far as I am aware, no permanent damage was done.
My initial work was helping with the development of radar for detecting submarine snorts or periscopes, which was a very pressing problem. I then moved on to its counterpart, i.e. radar mounted in the periscope of our own submarines. I moved on from that to working on Mark-5 IFF: Interrogation Friend of Foe. This was a joint Anglo-American project, run by a multinational Combined Research Group. I was responsible for developing a simulator, showing how, when a very large number of interrogators, ships and targets were involved, the system would cope with mutual interference technically, and how one could generate a suitable display under those circumstances. Designing the simulator was quite challenging, because I had to develop novel components – wideband synthetic delay lines, pulse transformers, blocking oscillator, and so on, which were all novel, as well as developing the circuit techniques themselves. This provided the basis for generations of more general tactical simulators, for many years into the future.
Thank you, Ralph. We have heard about your early years and moved on to your early career in the Scientific Civil Service at the Admiralty Surface Weapons Establishment. That was during the war. Now perhaps we could hear what you did after the war at ASWE.
During the war, all the people who had joined the Scientific Civil Service were there on a temporary basis. Therefore, immediately after the war, there was a so-called reconstruction competition, in which we were all evaluated to see what permanent job, if any, we could be offered. This was not just in our establishment but for the whole of the civil service. In this, I got 295 marks out of 300, which was the highest mark of anyone nationally. I was appointed, because of my age, as a Scientific Officer. We also had a postwar Radio-Location (i.e. radar) Convention, organized by the Institution of Electrical Engineers (IEE), in which I gave an invited paper on blocking oscillators. That paper was still quoted and used many years later. One of our immediate projects after the war was to apply the work we had done to the benefit of the civil world. Hence a small team was put together and given three months to design, build and test a prototype civil marine radar. We did this, and made the full design information available to all industry worldwide – not just in the UK. However, it was many, many, many years before any civil offering matched the performance of our prototype.
Following this project, and influenced very much by one of my naval officer colleagues, I concentrated on the problem of force-wide command and control information and intelligence collection and dissemination. We had quite ingenious manual methods which, however, simply were not good enough to deal with synchronized aircraft attacks. The Navy was also very worried about the problem of attackers coming in low below the radar horizon. One had to depend on information from outlying picket ships to defend the main units of a force. Some attempts had been made, both in the UK and America, to relay the video information from a picket by a ship-to-ship microwave link. This was quite difficult, because the antennas on both had to be kept pointing at each other whilst both ships were maneuvering and also rolling and pitching. In any case, it was limited to the ship-to-ship horizon range of, at best, about 20 miles but normally a great deal less than that. I felt this was a very inefficient way of using information, because all we were really concerned with was finding and transmitting the location of individual aircraft. I thought about efficient ways of encoding and transmission of information. This resulted in digital data links.
In this context I developed what is nowadays known as Information Theory. When in 1946 I went to the United States for trials of the IFF system in which I had been involved, I was at the same time asked to brief the US Navy on this digital data extraction and transmission system concept, and what we had done to implement it, which I'll speak about more fully in a moment. They immediately said, "You must go and repeat this to Bell Labs." When I told Bell Labs what we had been doing they said, "Do you mind repeating this? We'll just get Claude Shannon to listen." He did listen, but he didn't comment. What I found out in due course is that Shannon had done theoretical work on Information Theory a bit earlier and in a bit more depth than I. However, Bell Labs and the U.S. Navy were years behind us on how to apply it to a practical system and get it actually implemented.
I derived the Cartesian location of the echoes, by interlacing in the radar display a marker, controlled by a joystick. We used a joystick because at the time it was readily available, but in my paper proposing this, I had included what I called the roller ball, now known as a mouse, which I said was going to be more elegant and would be the long-term solution. One of my naval colleagues marked in the margin of my paper:, "The elegant ball-tracker stands by his aircraft direction display. He has one ball, which he holds in his hand, but his joystick has withered away." At the time all the plan position indicators for radar used mechanical rotation of a radial deflecting coil. However, in order to be able to interlace markers in any arbitrary position, we needed fixed coils with X and Y deflection systems. One of my Polish colleagues developed a very efficient and accurate way of doing this. But we needed multiple track markers, one for each target and therefore tracking operators had to align these in turn. In order to diminish their workload, I introduced rate aiding so that, from the pattern of consecutive displacements, the circuitry could deduce the velocity of a target and extrapolate it automatically. The operator had only to monitor whether the dots stayed aligned with the target, and make adjustments where they were not. These adjustments, again, corrected the velocity as well as the position.
We did not have digital computers at the time, so I had to develop a program-controlled time-shared analog computer. The principal method of computation was a switched-capacitor one, in which a capacitor built up a charge – in this case proportional to the velocity – which, in frequent finite increments, was added to the charge in another capacitor, representing the X or Y displacement.
This was at least twenty-five years before the switched- capacitor technique became reinvented and used more widely. As I already mentioned, both the technical approach for extracting information and displaying it, and the digital plot transmission concept, briefed to the Americans in 1946, including the mouse concept for correlating data between electronic storage and displays were patented in my name in 1947.
The target’s location was not displayed as a simple dot, but as a symbol, indicating friendly, hostile, neutral and so on. The user also needed to know whether a weapon had been assigned or not, a track number and so on. One could not normally put all of that directly on the plan display. Therefore the interlaced marker was used for a cross-reference to a subsidiary display which gave fuller amplifying information. The symbols on both the subsidiary and main display were drawn with the cathode-ray electron beam, using what is nowadays known as microprogramming – which again was novel at the time. In addition to handling direct radar tracks, we were also concerned with jamming. We therefore extracted the bearing lines of jammers. With digital plot transmission, we could link the bearings from most of the ships, to identify the location and tracks of the jammers.
The same time-shared switched-capacitor computer technique was used not only for target tracking but also for helping intercept officers to direct a fighter onto its target: We could use the known velocities to extrapolate where both the fighter and the target would be on the basis of the present course of the target and a postulated course for the fighter. We automatically allowed for the extra time taken in turning onto the course and turning onto the weapon launch course at end of intercept. This allowed one intercept officer (even an inexperienced one) to control two fighters at the same time.
Fighter direction had of course to be preceded by the decision which fighter or other weapon should be allocated to which target. The same computational technique was therefore also used for threat evaluation and weapon assignment. For this, we looked at all the targets and on the basis of their location and velocity, to decide how soon they could reach weapon release, and therefore which were the most urgent ones to engage. We then looked at all the weapons – fighters, guns or missiles – and ordered them in an order of priority, dependent on their flexibility. There were bound to be future requirements which we could not yet anticipate, and we wanted to handle each threat with that weapon which was adequate, but whose use would least restrict our freedom to deal with further requirements that might arise. If a short-range weapon or a fighter that was getting short of fuel would suffice, we did not want to use anything that had more range, capability or endurance. The general principle throughout the system was the use of electronic techniques to assist and support, but not to pre-empt human judgment.
Whilst we still carried on our own work, a contract was placed with an industrial firm to build concurrently two prototype demonstration systems, one for ourselves at the ASWE and the other to the US Naval Research Lab. Unfortunately, the firm made heavy use of electromechanical techniques, some adapted from telephone exchanges, which proved to be of rather limited reliability and flexibility. In the meantime, in my own lab, I continued to produce electronic techniques with a very much higher reliability. Altogether we put very high emphasis on getting reliable service. Each design proposal was reviewed by our own team, including the naval technical officer, at the stages of
• first proposal,
• first lab implementation,
• prototype paper design,
• prototype model,
to make sure it was going to be both reliable and easily maintained. We used modularity so that, if anything broke down, the faulty module could be easily identified and replaced. We used redundancy so that we had spare capacity. We used online monitoring facilities to show the status of things, and diagnostics that, in a sequence of predefined steps, reduced the uncertainty or where a fault might be located, down to the replacement module (and subsequently to the faulty component in that module). These were all new and revolutionary techniques at that time. Once the equipment was in service in the Navy, there were many occasions when the system had slightly less capacity than that for which it was designed. However there was no single occasion when it was not fully available for operational service – ever.
We also used techniques evolved from the simulator, previously produced for IFF, to simulate the radar, display, data-links and tactical-control system – not just for a ship but for the force – putting in real modules as they became available, and doing the rest by simulation, to test the design at each stage to see how it worked and how it interacted with the rest of the system and its operators. The result of this was not only that final acceptance trials had to take a very minimum of time, but also that they resulted in not a single design change suggested by the users, even as a wish list, and that we then had the same simulator to train the first operational crew. Therefore, we did not have the normal delay of many months between new equipment coming to a ship and the ship's crew being ready to operate it.
There were some other projects I know you want to tell us about, and then of course we will get to your later career at ASWE where you had a pretty fast promotion track.
The two systems I described were known as Comprehensive Display System (CDS, and Digital Plot Transmission (DPT). All our internal reports and designs were passed to our American colleagues at the same time of distributing them internally. The U.S. Navy was quite excited about the whole thing, but by the time they really got up to speed, digital components had become more widely available – especially in the States – and so their system, the Naval Technical Data System (NTDS) used the same concepts but more up-to-date technology. Digital Plot Transmission also needed adaptation to meet American requirements, so our original DPT became Link I and, as adapted in agreement with the U.S. Navy, it was called Link II. This was written as a Roman (rather than Arabic) two, and was subsequently interpreted as eleven. Link 11 is in fact still a British, American and NATO standard to this day.
Reverting to threat evaluation and weapon assignment, the Americans at that time did not have an automatic system for this as we had. In any case, they worked with a totally different approach and philosophy. We were concerned with making the best use of a limited number of weapons. The U.S. Navy was concerned with making sure that no weapon that was within range of a target remained idle – a simpler doctrine and one probably appropriate to their situation, but not to ours. However, this would not prevent operation in a joint force, with allocation of sectors of responsibility to each participant, and each working in their own sector according to their own doctrine.
There was no point in having a good display system without a good data source, and this required three-dimensional surveillance of the air space. At the time the accepted technique for doing that was the so-called V-beam: two fan beams, one vertical and one angled off, so that the delay between a rotating radar passing the target on the vertical and the inclined beam give an indication of height. This was severely limited, both because of its vulnerability to jamming and because, with multiple target, it gave confusing and ambiguous information. The obvious step to avoid this was to have a vertical stack of several narrow beams. However, this also had limitations, both in the sheer cost of using so many transmitters and receivers and in that height could only be estimated to an accuracy of maybe a quarter of the beam separation. My object was to produce a smaller number of beams, which is scanned in the vertical direction. In order to avoid discontinuity where one scan finishes another one starts, the scans were tilted, so that the top of one scan was laterally displaced from the bottom of the one above. Combining this stack of incremental sector-scans with the rotation of the radar itself, the composite scanning angle was 45°. The vertical and horizontal projections of this scan across a target could be displayed separately, to derive the target’s height and its accurate azimuthal position.
This set of scanners, stacked on top of each other, could not have been put in front of a reflector mirror, as it would be too much of an obstruction. Therefore we used a microwave lens. Since the beam from a source is diverging in a spherical wave front, and we want to form a plane wave front, this lens was designed to convert from one to the other. To do this, we put in waveguide sections of somewhat different widths, so as to have a variable refractive index, to permit a smooth circular wave front at one end of the lens and a smooth plane wave front on the other. The whole system of source and lens and casing for it all looked like a large torch. It was mounted on trunnions, so that the stack of scanners could be kept vertical as a ship rolled and pitched.
One of our big concerns was vulnerability to jamming. We may have been very well placed on this, because ASWE was the world leader in jamming techniques. We had all the knowledge of what jamming could do, and the cross-fertilization between a poacher and a gamekeeper. Of course, the narrow beams are an anti-jamming feature in themselves but in addition, each beam was on a different frequency. When we found that the scanner produced a mismatch to the magnetron transmitter, varying over the scan rather than trying to correct this, we deliberately maximized and exploited the effect in order to get pulse-to-pulse frequency changes, and applied the same frequency changes to the local oscillators, so that the receiver would work consistently. We also had pulse-to-pulse changes in pulse repetition interval, in order to foil deceptive jamming using artificially delayed pulses. Again, these anti-jamming techniques were probably about a decade ahead of what was practiced elsewhere. For the radar, I had the leading part in developing design and concept, but was not responsible for its implementation – unlike the display system where I was responsible for the implementation as well.
In aircraft carrier operation, one was also concerned with minimizing the time when a carrier has to head into wind, for the recovery of returning aircraft. Therefore I developed a scheme, included within CDS, in which all the returning aircraft were put onto consecutive members of a set of concentric range rings, centered on the input end of the approach path to the flight deck, shrinking at the approach speed, and spaced by the permissible interval between aircraft landings.
The installation of the combined system of systems in a carrier was first completed in 1957 in HMS Victorious. Benefiting from the fact that the crew could be pre-trained in the simulator, she was able to go almost immediately to the States for trials against the U.S. Navy, in which Victorious spectacularly lived up to its name. We were so successful in intercepting all the raiders that the U.S. Navy sent against Victorious, that the USN believed we had more aircraft on combat air patrol than Victorious had altogether. In fact, we did not have any aircraft any combat air patrol, but sent them up from the deck, as required. The USN also assumed we must have very superior airborne intercept radar on our fighters when we had none at all. Furthermore, they assumed we had highly trained aircraft direction officers when ion fact they were brand new, and had had only trained on our simulator. However,, with the benefit of our computer, each could handle two concurrent intercepts.
To save your blushes, perhaps I will actually quote from the book Cold War, Hot Science which is about the UK's defense research from 1945 to 1990, and is sponsored by the Science Museum and the Defense Evaluation & Research
Agency of the Ministry of Defense. It concludes, "Ralph Benjamin and his colleagues were responsible for some of the fundamental technologies of the second half of the 20th century." That's a lovely quote. Now we come to your later career at ASWE.
In 1954, on my 32nd birthday, I was promoted to Senior Principal Scientific Officer, because this was the youngest age at which it could be done. In fact they backdated it to an earlier starting point. I was made head of the Command and Control Information System Research Group, and five years later on my 37th birthday, again backdated, I was made Deputy Chief Scientific Officer and Head of Research & Assessment. As such I had to deal with all the work on radar, communications, guided missiles, jamming and so on. I had a very long practical working day normally taking work home and not finishing until about ten in the evening. I then spent late nights and week-ends on trying to generate a common theoretical basis for thinking about all our projects. I did this in the first nine months of the new job. It was also accepted as by London University a Ph.D. thesis, and was published as a textbook by Pergamon Press. Much later I found the book had also been translated into Russian, where it had a much bigger distribution than it ever had in the western world.
One of the projects I initiated, but had no involvement in its detailed implementation was, the first operational ship-borne satellite communications terminal. It worked with the Early Bird satellite, using an antenna which was automatically kept trained on the satellite, as the ship moved relative to it.
Another project was in communication, where the naval doctrine was that every user of communications must have his own transmitter, receiver, antenna and frequency channel, accepting the severe limit on the number of antennas and frequency channels available within the confines of a ship. Therefore, against a good deal of opposition, I caused a system to be developed, in which the optimum number of channels that could be efficiently accommodated on a ship was installed, and allocated dynamically to those who instantaneously had need for it.
There was much excitement at the time about ship-borne-guided weapons. It was proposed that the guns should be got rid of altogether. Although I liked modern technology I did not think this was a very good idea. I suggested that if one of HM ships had to stop a lot of sampans running illegal guns into Malaya or dhowns running illegal immigrants into Hong Kong, using one of the few (~ twenty) expensive guided missiles kept aboard a ship was not the best way of doing things. The counter-argument was that the top-weight as well as space taken by a gun was an excessive burden on ship design, and that the manpower was difficult to accommodate. I accepted the challenge to produce a gun that had far less top-weight than its predecessors and needed minimal manpower. This was the 4.5 Mark VIII, which in due course made an essential contribution to the Falklands War, and is still fitted.
We have heard quite a lot about your career and your rise through ASWE. Then you changed establishments and went to the Admiralty Underwater Weapons Establishment (AUWE). Perhaps we can hear something about the nature of that establishment and what you did there.
In '64, when I was 41 years old, I was made Chief Scientist of the Admiralty Underwater Weapons Establishment. This had been formed very recently by the merger of one establishment concerned with torpedo technology coming from Scotland – one with mining and counter-mining, coming from Havant, near Portsmouth, and underwater detection, i.e. sonar, at Portland. All these were brought together at Portland, and we also had a research group, which combined elements from all these and some from the Admiralty Research Lab at Teddington. I had four departmental heads, who individually were very good but barely on speaking terms with one another. Soon after taking over, my responsibilities were enlarged from Chief Scientist to also being the Director of the Underwater Weapons Establishment, a position previously was reserved for a naval officer, and also being the Ministry of Defense (MoD) Headquarters Director of Underwater Weapons Projects and a separate Directorship of Underwater Systems Research. The staff I had was about 2,500 civilians and fifty naval officers including three captains and a couple of RAF office, who were concerned with sonobuoys and air-dropped torpedoes.
I was younger than any of my four deputies, as well as the division heads and naval captains. Fortunately I managed to establish a good relation with all quite quickly, but there was quite a big problem in integrating the establishment and generating internal and external cross-fertilization. In regard to research, I reallocated the research in the torpedo, sonar and mine warfare areas and weapons areas to the respective departmental heads. I wanted to maximize cross-fertilization between research and projects both ways – so that the research people could understand the operational needs and the projects people could understand what research outputs could do for them. I believed that this was worth any small loss in cross-fertilization between research activities in different applications areas To further mutual simulation, I arranged brainstorming sessions whenever I visited any section, and more formal project reviews, in which people presented their achievements, problems, queries or results to all but the most junior scientific staff. I encouraged people to let their hair down and throw in ideas, as I did myself. If an answer was "yes but," I felt I had done something useful in stimulating a new way of thinking.
I also encouraged interaction with outside organizations. The underwater weapons people had traditionally been a closed community, having a different environment, different units of measurement and different ways of doing things and thinking about things from everyone else. This interchange was by no means one-sided. The ocean is a very difficult environment. The low propagation speed of sonar is much more difficult to separate from a ship's speed, than distinguishing radar propagation from aircraft speeds, and sonar echoing delay times are too slow to permit sequential scanning over 360º. Sonar transducers have to generate pressures that are comparable to that of the surrounding water, which can produce cavitation, preventing sound transmission. As a result, the sonar community had long lived with the concepts of pulse compression, spread-spectrum modulation, phased arrays; and combined range and Doppler processing, which the radar community had to rediscover, in dealing with intercontinental ballistic missile, and sonar people ;long had to cope with multi-path propagation, a problems now critical to broadband communications.
Other areas in which we had to do more than perhaps was common in other spheres were in the analysis and exploitations of hostile emissions, the development of sophisticated homing logic in torpedoes and countermeasures against it, and devising highly intelligent mine fuses and ways of trying to combat them. Whilst the underwater weapons area thus often had more advanced system concepts than my old colleagues in the radar world, I discovered that the technologies for meeting those concepts tended to be more conventional and limited.
Torpedoes, when operating against submarines, had to cope with the high pressure of great depths. In designing for this, we had to understand and cope with the problem of buckling stresses. This is something that civil engineering had not yet learned, as demonstrated by some spectacular collapses of early box girder bridges. We also had to deal with metallurgy, with corrosion, electrochemistry for batteries, hydrodynamics, aerodynamics for air-transit antisubmarine weapons, etc. There were very sophisticated engine concept, for operating with no access to external oxygen, and when it is important that exhaust gases do not give away where a torpedo is operating.
Torpedoes were connected by guidance wires to their parent submarine, and the optimum combination of the information available from a torpedo’s own sensor system and the submarine sensors, changing all the time during approach, posed interesting problems. A submarine and a torpedo are primarily weapons of stealth. Keeping them stealthy was a considerable challenge. In addition, sophisticated but very different fuses were required for a torpedo used against a submarine and one used against a ship. For the latter, the aim was to pass underneath the keel and explode there, so as to break a ship's back. Against a submarine, the aim was to lock onto a vulnerable part of the hull and achieve hull penetration. Mines had a sophisticated fuse of their own, which I may mention later. We had the technologies of sonar transducers, very sophisticated signal processing of the phased array, combined range and Doppler, and so on.
Since submarines are concerned with stealth, keeping quiet and listening to its target’s emissions, the suppression of noise radiated by a submarine or torpedo, and the suppression of noise radiated by a surface ship were major problems, in defense as well as offense. Sonar performance is limited by noise picked up by its transducers, so the reduction of noise generated by our own propulsion system or hydrodynamics also became tricky problems. Sonar dunked from a helicopter has unique problems of its own in operation and in also in recovery. This is because as it is withdrawn from the water it will normally have some velocity relative to the helicopter, and the conservation of angular momentum means that, as the cable is shortened, the oscillations get more and more violent. Unless appropriate precautions are taken, the sonar cable will wrap itself right round the helicopter. We also had the usual problem that anything going into aircraft has to meet criteria of airworthiness. This can add substantially to the cost of the time in getting anything into service. Much was dependent on the ocean environment, and so oceanography was another big commitment.
One form of detector of submerged submarines was Magnetic Airborne Detection (MAD), which despite its name was not actually mad. However, it had to take into account, both the temporal and spatial variations of the Earth's own geomagnetic field. To protect submarines and surface ships magnetic mines, these vessels needed degaussing, using suitable electric circuits to cancel their magnetic signatures. This was far from trivial, because the magnetic signature depends on the vessel’s location and orientation in the Earth's magnetic field, and its history of magnetic exposures.
We had a very active operational research group, to develop the concepts for the optimum deployment of ship sensors and weapons to use in antisubmarine warfare, submarine operations, and mine and anti-mine warfare. One of our operational researchers was seconded to the Far East Fleet at the time of the so-called Indonesian confrontation, when one of the big problems was the smuggling of arms from Borneo into Malaya. The small number of relatively slow patrol craft available proved very ineffective against this. He used common sense and operational research to decide possible starting points, destinations and schedules for arms smugglers, wishing to do all their transit during the night. So he identified the times and places targets were most likely to be found. and concentrated all the patrol effort this area. This produced quite a spectacular improvement in interception.
In regard to mine warfare, the mechanical sweep of the First World War against moored mines was irrelevant, because a gadget had been devised which allowed a sweep wire could pass through the mooring wire without disruption. There is no time to explain how it works, but it is really quite simple. Therefore sweeping tended to try to simulate the signature of a ship, for which a mine fuse would look. This could be a combination of magnetic, acoustic and pressure signatures. Our sweeps became very clever at simulating these signatures but, even then, they could be fooled by intelligent mine fuse, which could look for a very specific type of ship, could be active only intermittently, or could be active only after having been triggered a certain number of times. Therefore the emphasis has shifted from minesweeping to mine hunting, i.e. looking for mines by ultra-high resolution sonar. Even this can be difficult, when the nature of the ground allows mines to be shielded or buried, or when there are natural or discarded objects on the sea bottom which may look similar to mines. Therefore part of the work is to survey the ground, in order to find where mine hunting is most suitable and to route ships accordingly. The dual- role of minesweeping and mine hunting vessel itself must of course have very low acoustic magnetic and pressure signatures so that it does set off mines. We were also responsible for diving technology for the three services and of course also had to use divers in experimental roles.
I know you want to tell us about assorted issues connected with AUWE.
I never like asking people to do things that I am not willing to do myself, or to interact with people about problems I do not understand. For both of these reasons I became qualified as a naval diving officer, and took part in some interesting trials. However, I could not take part in the really enjoyable ones overseas.
In the underwater world, our work was very specialized. There was very little capability or understanding in industry or academy. In addition, few allies had anything like the background on underwater warfare that the Royal Navy had. This probably was one of the reasons why the Parliamentary Select Committee on Science & Technology, when it was first formed, made us the first government establishment to visit. I believe they found the visit well worthwhile.
Soon after I joined AUWE I was offered to leave the public service and to become the Head of the Department of Electronic Engineering at Birmingham University. However I felt that my work at AUWE was more important. Indeed I had many similar offers from industry and academia, in the UK and abroad.
One of the peculiar administrative problems we had at Portland was that the mess waitresses all disappeared in the summer, when they could get better jobs at the Weymouth seafront. I overcame this problem by enrolling elderly motherly ladies, who loved to look after us, but who couldn't get jobs commercially because they were no good at arithmetic. I therefore got our staff to work out their own bills.
We had an active program for training apprentices, which benefited the wider world at least as much as ourselves. The training was very old-fashioned, so I started forming the apprentices into team, to design and build go-carts and racing them against each other, which revitalized it quite considerably. In order to stimulate interest in and identification with the establishment, I felt we ought to have an open day. This was quite a difficult balance: Planning and preparing for it was a disruption to work. However, at the same time, it was an incentive to get things done in time to be able to show them, and in the longer term it did a great deal for staff morale, and I believe it was well worth doing.
I did much to get to know and to support my own staff. Our lab mechanics were able to work closely with scientists, to understand almost any job, and to use almost any machine tool, and went with us aboard ship, to support experiments at sea. With difficulty, I therefore negotiated that they got more pay than normal technicians. This was not popular with the trade union nationally, but of course it was popular with our own local shop stewards. At meetings with the staff
associations, I tried to see their point of view and be as helpful as I could without prejudicing official policy. The result was, when there was a pay strike in which all the defense establishments were paralyzed, we kept on working normally. I got a lot of headaches when senior naval officers visited industrial firms. Very often they got conned into making commitments on behalf of the Navy which were very damaging to us. It was difficult to find a way around that.
The dramatic advance in the capability of Russian submarines had a very big impact on Royal Navy thinking. One result of this was that I inherited from my predecessors a “crash program” for producing an advanced anti-submarine torpedo, in which development, tests and more particularly trials were to be compressed in a way that my staff regarded as wholly unrealistic – as I did when I took over. I had the really big headache of stopping the crash program from living up to its name and crashing. The anti-submarine torpedo finished up being successful, and also having a very powerful anti-surfaceship capability. However, this upset some of the key people at the Ministry of Defense, because they were very keen on getting the modern image of airborne-guided weapons against surface ships, and were afraid – probably correctly – that funding of that program would be prejudiced if it were known that we had a torpedo which was very effective against such ships. Therefore they suppressed information on the capability of this torpedo. Other problems arose when we discovered the principal contractor for torpedoes charged the MoD a very high cost for testing and repairing faults found in manufacture. Investigation proved that there was a scam, worked jointly by the workforce and the management, in which the testers sent a large number of perfectly good weapons back for rework. They got extra piecework pay and the company got extra pay for work that either was not done at all, or did not need doing.
The combination of looking after a big establishment dealing with a lot of these frustrating and non-scientific issues, while trying to be actively involved with the scientific staff down to lower working levels made for a very busy life. I had to do a lot of commuting between our own sites and between our main base at Portland and the Admiralty offices in London. To get to London in time, I needed a car to Salisbury to catch a train there. Therefore, I had a light installed in my official car so I could work during my travels. Our establishment was also the biggest employer in the relatively small town of Weymouth, which meant I had to be involved in quite a lot of official local activities, including being governor of the local technical college.
Appointments at my level were normally for three and a maximum of five years, but I was asked to stay on for seven because of the critical importance then attached to antisubmarine warfare and, I was told, because of my crucial role in this. However, after seven years, I did move on. They arranged a traditional naval departure ceremony, with. my car s towed by a team of naval captains and commanders. It was a nostalgic occasion for me. I had an acknowledgement by Sir Frank Cooper, the Civil Service Head of the Ministry of Defense. He wrote:
"On behalf of all your colleagues at the Ministry of Defense I express our appreciation of your highly valued services. From the time you joined in 1944, you established a reputation for original thinking and depth of theoretical insight, which became your hallmark. It was recognized by your many patents and your unusually early promotion at all levels. Your pioneer contributions to radar and signal processing and to action data automation laid the foundations which continue to be built upon to this day. Your ability to lead and inspire research teams led to your becoming Deputy Chief Scientist ASWE and the Director AUWE."
The letter closed: "You did much to enrich the science of underwater systems and even more importantly to raise the level of professional capability of the establishment. It is with considerable regret on its own behalf that MoD sees your transfer, in the national interest, to the Government Communications HQ."
Thank you, Ralph. You have been good enough to tell us about your career at AUWE. Now perhaps you could go back a bit and tell us about the Navy in general and the MoD, and some of the external activities in which you were involved. That would be most useful.
After twenty-seven years working in and with the Navy, it was quite a break to lose direct contact – although, I am happy to say, I continued to have some involvement with the Navy ever since. One of the remarkable things about Naval officers, or any service officers, is that at sea or in war they have to take quick decisions above all, whereas, in other posts, they must take their time, if necessary, to ensure that they take the right decisions. They have three-year appointments of which a significant part at the beginning goes into learning about the new job, and some at the end in preparing for the next post. During this time, they are expected to formulate policies or direct projects that take ten years or more in implementation. There is then a tendency to work like a dinghy helmsman steering a supertanker, with frequently changes of rudder: direction does not change much, but it impedes progress. Nevertheless, I was very impressed how many Naval officers managed to be very effective in the two very different sorts of environments.
A significant difference between the Royal Navy and the U.SN is that the Royal Navy has specialist officers in their own dedicated branches. All USN officers are generalists. That means that, at the lower ranks, USN officers depends much more upon Warrant Officers to provide the specialist skills, but in the higher ranks they have got a wider perspective, which makes it easier for them to see a broad background for strategic policy decisions. Our culture was one in which officers gave their loyalty, in order of decreasing importance: first to the ship; then to their specialist branch of the service; then to the service as a whole; and then to the country. This did generate good bonding. However, it could go counter to effective and balanced decisions. I had quite a lot of contact, then and later, with noncommissioned officers of all the services, as well as with officers at all levels. I was always impressed by the quality of all of them. However, I did feel that people enlisting in the ranks, of any of the services, tended to do it because having a uniform made them feel big and grown up and important, when in fact they were in a sheltered environment, where most decisions were made for them, and the real need to face adult responsibilities was delayed.
Turning to the Ministry of Defense, in my time at least, there were a number of defects. I am not giving a balanced picture here. I want to emphasize all the things that could be improved. One of them was that any technical or procurement problem had to undergo a very long and complex approval process, so that, by the time a project actually started, both the concept and the technology, was already becoming obsolete. Furthermore, some of the staff originally involved with formulating it, and geared to take it to success, had already changed to new appointments. The very great care and hence time taken to try to make sure that the right decision was taken, in fact made sure that what was eventually done was less cost effective, and had less time before being overtaken by obsolescence..
That latency problem still exists today to a great extent, doesn't it?
I fear so, yes. In addition, managers got great credit for being on time, on budget. That was an incentive to put hidden contingencies into the plan, and to adopt low-risk but also low-gain solutions. Another source of inefficiency is that if, in one year, a Department or project spent less than budgeted, the conclusion was drawn that it did not need so much money and hence it was cut for the following year as well. This obviously provided a powerful incentive to spend up to budget, come what may. Treasury rules meant that funds were released for only one year at a time. Therefore firms could never be sure whether they would get a continuation contract, and procurements had to be in drips and drabs, rather than by bulk orders. Therefore the taxpayer lost much in the way of economies of scale.
Multinational projects were popular, as they are now, not least because the political commitment to other countries meant it would be difficult, at a future time, to cut a project. However, difficult interfaces, different standards, linguistics, distances and different policies meant it could be very hard to optimize such a project. I felt t that, in the majority of cases, it would have been h better if one could agree to share the responsibility of lead nation for different projects between countries, letting the lead nation be the sole manager of the project, free and indeed encouraged to place subcontracts where-ever that was the most effective way of doing the job.
In the Ministry of Defense, as in NATO later on, I found that some other people were more effective than I in using engineering principles to manipulate the system to reach a desired decision. One had to modify the desired output by the inverse of the transfer function of the political and approval process, and use this as the input to the system.
In addition to my direct responsibilities at ASWE and AUWE, I was involved in quite a variety of external activities. One was as chairman of committees of the so-called CVD, the Committee for Valve Development, (“Valve” here stands for electronic tube.) Others were the interdepartmental Advanced Computer Techniques Steering Committee and the predecessor of what is nowadays known as the Defense Scientific Advisory Council (DSAC). I was also a very active member of a very small Air Defense Study Group, reporting to the Board of Admiralty, and of the Chiefs of Staff Air Defense Working Party. NATO-wise, there were committees for studying the application of science and engineering to defense. They set up two intensive working parties, one on Man & Machine, and one on Command & Control, largely at the two-star or three-star level. In both of these the chairman was an American, and I was the deputy chairman, but also, in both cases, the chairman was there for the opening session and then returned to the States. During the two or three weeks of intensive work we had in Paris, I was in fact acting as the chairman.
There were a large number of European bilateral collaboration projects and the so-called Tripartite Technical Cooperation Project between the UK, USA and Canada, and in practice also Australia and New Zealand. I had an interesting experience at a NATO conference on antisubmarine warfare in Malta, where the leader of the German delegation was Admiral Kretschmer, and one of the officers on my staff was the commander who had captured him and his U-boat during the war. They got on very well with each other. I was also co-opted by the United States for Project Charles to plan the air defense of the North American continent. They were keen to retain me for its following project Lincoln, but MoD decided they could not spare me. I also had a lot of fruitful contact with industry and universities, including acting as external Ph.D. supervisor for, amongst others, research students working under Professor Dennis Gabor at Imperial College, the Nobel Prize winner for his work on holography.
In visits to government and industrial labs, I always interacted very fully with the host, with comments and constructive suggestions. Probably as a result of this, I was always sought and welcomed as a visitor, particularly in States. They normally organized all-day sessions I starting early, sometimes at breakfast, usually followed by evening social events. The penalty of this was that, when I got back to my hotel, which often was not much before midnight, I then had to take a couple of hours to write notes to record the information I had received, for the benefit of my colleagues, as well as noting down any ideas of mine, stimulated as a result of the interaction. Of course in disseminating these notes I had to be very careful to preserve any confidentiality. Similarly, in my reactions to the hosts, I had to be careful not to give away anyone else's information. My wife, Kathleen and I always tried to return any hospitality I had received when people from abroad or from industry visited us. It could be quite a burden. Unlike my naval colleagues and naval subordinates, we never had an entertainment allowances, or naval stewards to help, although later at GCHQ we got refunds for some entertainment costs.
Now we come to the next step in your career, which was the Government Communications Headquarters (GCHQ) at Cheltenham, UK. Perhaps you will be good enough to tell us what you are allowed about your work at that establishment.
GCHQ is concerned primarily with the collection of Signals Intelligence. It is the equivalent of the American National Security Agency (NSA). Before I arrived there, I had very little contact with Intelligence, other than occasionally interpreting some information for the Department of Scientific and Technical Intelligence in the Ministry of Defense. One of my senior members at Portland had developed a growth in his ear, which made him very deaf. In addition to being a Ph.D. and a very good scientist, he also happened to be very interested in languages. After he became deaf, he became a fluent lip reader not only in English but also in other languages, including Russian. Partly as a result of this, I recommended him as Scientific Attaché to Moscow. By all accounts, his ability to listen by lip reading to things that he otherwise could not have heard proved very valuable. Unfortunately, when he finished his term there, he was not allowed to return to us, because it was s assumed that anyone who had had a lengthy appointment in Russia could have become compromised.
Our function at GCHQ was the collection, analysis, interpretation and possibly verification of Signals Intelligence. Signals Intelligence is a combination of the analysis of ELINT: electronic intelligence – that is the emissions of technical equipment – and Communications Intelligence: COMINT. First of all, Communications Intelligence does traffic analysis, which works out who is talking to whom, and deduces the order of battle, the organizational structure of forces and the level of activity. The second thing COMINT does is cryptanalysis, which endeavors to decrypt coded messages, in order to get detailed content, from which one can deduce capability, vulnerability, plans and activities. Some of the signals intelligence work merely provides long-term background, but much of it is a real-time battle of wits. We also had the complementary function of ensuring the security of our own communications, COMSEC. Apart from being an important function in its own right, this provided cross-fertilization: the gamekeeper of COMSEC could teach the poacher and a poacher of SIGINT could teach the gamekeeper. COMSEC, the secure coding of our own communications, is more demanding in peace than it would be in war. This is because in peace a potential enemy can take months, or even years, to deduce information that would help him in the case of future hostilities. In war, information becomes irrelevant as it gets overtaken by operational developments, normally in a matter of hours or at least days.
A great strength of GCHQ was that its administrative staffs were people who were fully aware of our function. Most had worked in Signals Intelligence collection or analysis and expected to work in it again. Therefore they quite understood that their role was not the enforcement of arbitrary rules, but to act as a support and enabler of the operational activity. Everyone was aware that we were engaged in an ongoing battle of wits with opponents, and proud of the fact that we did rather well in it. Therefore we had very high morale and motivation. Much was made in the media of problems in regard to trade union recognition. In fact, the national civil service unions were not very effective in involving people whose primary loyalty was to a well-motivated organization. The initiative to form their own staff association, although clumsily organized from Whitehall, had overwhelming support of the staff, although there were some well-publicized dissenting voices.
What was my personal role in GCHQ? My predecessor was responsible for coordinating research and development on SIGINT collection. I inherited that function, but I also had a much wider role, not only at GCHQ but also as Chief Scientific Advisor to the Security Service and the Secret Intelligence Service, and as Coordinator and sponsor of research and development for all the intelligence services. I was also Head of the scientific and engineering profession for the Foreign and Commonwealth Office. Within GCHQ, in addition to SIGINT-collection R&D, I was responsible for scientific aspects of cryptanalysis and of intelligence collection and interpretation, for operational research, and COMSEC R&D. SIGINT engineering involved procurement, installation, maintenance as well as development. We had the management of our own communications net, the acquisition and operation of computers, as well as the development of our own special-purpose computers, and computer services which we provided to other government departments. I also directed the cross-governmental Joint Speech Research Unit, where, amongst other things, we did pioneer work on speaker recognition and speech recognition. We also had a cross-departmental high-frequency radio advisory panel, and we did pioneer work on (radio) direction finding and what is now known as super resolution.
I was visiting professor for the University of Surrey, and one of the things I sponsored there was the evolution of the use of cheap commercial components with redundancy and reconfigurability, in order to provide high availability, for use in satellites. This resulted in the development of cheap satellites in a separate company, which is still very active and successful, internationally. Although our own work was highly classified, the scientific principles behind it were not. I continued to be personally active in work with Learned Societies, in personal publishing, and university guest lecturing. One man at GCHQ, responsible for the analysis of the capability and design of Soviet block intercontinental missiles, had been trained as a linguist, but had become a very competent electronic and general engineer through the study of Russian textbooks. One of the books he had studied, and according to him, the one he found most valuable, was the Russian translation of my book “Modulation, Resolution and Signal Processing”.
Shortly after I arrived at GCHQ, the post of Director of the whole Royal Radar Establishment at Malvern became vacant due to a sudden death. I was then put under considerable pressure by the Ministry of Defense to leave GCHQ to take over that job. However I refused this, partly because I thought it was unfair to GCHQ to have another disruption when I had just arrived, and also because I felt that the job at GCHQ, though less esteemed in the public eye because less known, was considerably more challenging.
Would you like to tell us about some personnel issues around GCHQ.
Yes. GCHQ was always heavily over-subscribed by applicants for appointments. That applied above all to mathematicians. It was well known that the most challenging work in the country for that profession was at GCHQ. Almost all our staff were lifelong professionals. Possibly in response to the difficulty that they couldn't talk to their friends and relatives about their work, they had an amazing variety of interesting outside activities and hobbies. One of the reasons why Cheltenham, a city that is not very big and in which GCHQ is located, has such a very rich cultural and sporting life is because of the outside activities of people from GCHQ. It also explains why Cheltenham clubs did so well in chess and Bridge, which are natural interests for mathematicians.
The professional classes we had were Mathematicians, Electronic scientists, Engineers, and SIGINT specialists. Only the SIGINT specialists came in the so-called administrative class of the civil service. They regarded themselves, with some justification, as an elite of administrators. When I arrived, there it was taken for granted that virtually any jobs with administrative responsibility should be filled by these people. I challenged this strongly, because I believe that each job should be filled by the person best fitted to do it. He should have all authority relevant to his post, and everyone's individual career should be guided to fulfill their fullest potential. A change of this sort could not be implemented without pain, because administrators had entered the service with a career expectation that would be undermined by this change. However I managed to move things fairly rapidly in this direction.
In addition to my quite wide and diverse administrative responsibilities, I personally was probably more scientifically productive in this post than in almost any other. I cannot say much about that, but I can touch upon a few of my initiatives. The interpretation of signals intelligence involves associating several separate items of intelligence to build up a mosaic. We must recognize when the same piece of information has come from the same source but received through different routes, so that it is merely duplication, and, when it is genuinely from different source, so that it is true validation. When a tentative hypothesis is formed, one can ask, "If this is really true, what else should be happening?" One can look for that with hypothesis testing. I formulated these principles more explicitly than I believe had been done before, and I also helped to develop computer techniques to implement these processes.
One really serious problem is that if one looks at a number of major political developments in the last few decades – like the Six Day War, the invasion of Czechoslovakia, the invasions of Hungary and Afghanistan and many others – in all there was very good Signals Intelligence that this was going to happen. In each case politicians said, "If I were in the shoes of the politician in the other country involved, I would not be doing this. Therefore it cannot be true." I urged people at the Cabinet-Office level to form hypotheses, and agree in advance what generic warning or indicators must set off an alarm, and what discriminators or means of elimination or confirmation should then be invoked to support the eventual judgment. If these analyses of hypotheses had been performed, and agreed jointly with politicians, at leisure and without the emotional pressure of events, it should be easier to reach the right judgments in a crisis.
It seems that politicians cannot think "outside the box" as they say.
Yes, but one can try to fit the right box. It has been the general doctrine that, if there is a crisis, mobilization or other pre-emptive action must be taken as secretly as possible. I suggested that, in most circumstances, this is exactly the reverse of what should be done. This is because, as long as the opponent is not publicly committed to war, if he sees that his intentions are being suspected and effective preparations are taken to counter them, he may still pull back. Therefore it may be a good idea to make sure that he knows that we are responding and preparing, rather than to keep it secret.
We produced a speech security system that was sufficiently robust that it could work over a wide variety of mixed-quality commercial channels. This was so effective that it was adopted by both Margaret Thatcher and the President Regan, for communications both between them and to others. On the other hand, I felt that it was very likely that the IRA might be listening to the communications from security services in Northern Ireland. They would not have the sophistication of cryptanalysis of a major nation, and in any case the communications were normally critical only over a matter of hours. Therefore the need there was to produce a cheap, lightweight and simple encryption system,. This we did, and it was also made available to the police in the UK mainland, because there also it was not unknown for criminals to listen to police communications. Our forty-two chief constables are very independent, and some had somewhat idiosyncratic ideas. Most accepted our encryption device with enthusiasm, but others said that it was good for the public to be able to listen in, and to know that the police are active on their behalf: if criminals occasionally benefited by it as well, so be it! When I paid a visit to Northern Ireland (in another context), I had to be escorted by a couple of corporals in civilian clothes but armed, and they took great care never to return by the route in which they had gone out. They explained that this was because although the vehicles used were unmarked, they were known to the IRA and, the moment we passed their observation points, a boy on a bicycle would go to the nearest telephone and, if we came back the same way, an ambush would be waiting.
I had quite a few occasions to brief senior politicians, diplomats and military people on relevant aspects of the work at GCHQ. In this I tried to bear in mind the words by Rutherford when he was Director of the Cavendish Laboratory at Cambridge: "If you really understand something, you should be able to explain the principles behind it to a barmaid – in other words a sympathetic listener with common sense and no preconceived ideas.” One of the people I had to brief was Peter Jay, when he was appointed Ambassador to Washington. I was told to expect someone who, in relation to our work, was totally ignorant but highly intelligent. I think this guidance was borne out. The most interesting briefing was probably that of Margaret Thatcher. I found very rapidly that she had a very quick and active mind. As soon as one told her something, she would try to jump ahead and decide in her own mind to what it led. Unfortunately, not having the relevant background, she was likely to jump to the wrong conclusion. It was then very hard to shift her from that. However, if one provided her with information and ideas sufficiently rapidly that she could only just keep up, this was a very effective and fruitful way of communicating.
Speaking of Margaret Thatcher, I have always felt that the significance of her decision to go to war over the Falklands has been almost universally underestimated – including possibly by herself. From our SIGINT work it was quite clear, as from much information in the open domain, that the Soviets believed the West was weak, effete and indecisive. I believe that Margaret Thatcher's decision to go to war rapidly, on behalf of 1500 farmers who were almost at the other end of the world, added more to the credibility of the deterrent than the Polaris and Poseidon systems, which we at AUWE had done much to incorporate in our ballistic-missile submarines.
SIGINT collection and processing is a worldwide (or at least alliance-wide) activity, and it depends much on inter-allied cooperation. We had bilateral links with many other countries, but by far the closest cooperation was, not surprisingly, with the United States: with the National Security Agency. Incidentally, coordination with them led to the one occasion when I crossed the Atlantic by Concorde. I left on a February evening somewhat after dark, and arrived in Washington at 3 o'clock the same afternoon, and suffered from no real jetlag. Most of my fellow passengers were the captains of aircraft from other airlines, who drank the free alcohol on Concorde nonstop for the whole flight. I presume they did not have other flights immediately after their arrival.
I made quite a lot of visits to the States and of course also had many visits from there. Indeed, we had wholehearted cooperation from both sides. As a result of my personal share in this, NSA wanted to recommend me for the highest presidential civilian award. This was vetoed by the Foreign Office, because at the time our relation with NSA was supposed to be secret. As a result of this, NSA decided to mint a special medal and present it to me at a big ceremony in their central lecture theater. The medal was inscribed, "Dr. Ralph Benjamin, with admiration and respect from the Director and Staff of the National Security Agency." It is also interesting that Admiral Bobby Inman, who became U.S. Secretary of Defense, was a former director of NSA and Dr. William Perry who became the Deputy Secretary of Defense for Research and Engineering and later Secretary of Defense, was a former chairman of the NSA External Advisory Board. I had been on first-name terms with both of them for a number of years.
When I reached my retirement age of sixty, the overall director of GCHQ, Sir Brian Tovey, recorded lengthy interviews with me, somewhat similar to the ones that Professor Hill is now conducting for the IEEE, except they were specifically aimed to record my ideas on how SIGINT in general, and not just SIGINT R&D, should progress. This was intended for the guidance of current and future members of the GCHQ Directorate, and for the orientation of high-flying recruits. He clearly regarded this as very valuable. How it has been used afterwards I do not know.
Now we have heard, Ralph, that you retired from your UK post GCHQ Cheltenham. Then came something completely different, which was your move to NATO. Perhaps we can hear something about that.
Yes. I did not feel ready to stop doing interesting and creative work. But in the UK there was a rule, applied to senior civilian Government employees - but not only to military staff - that after retirement one could not take a job in industry
in one's own field of work for at least two years. This is to avoid any suspicion of corruption or favoring one firm over its competitors. However, it results in the waste of a lot of talent and knowledge. It is often bypassed by a variety of ruses, but I was not keen on that. Therefore instead I applied to NATO, for a so-called freelance job. These differ from the top posts in NATO, which are reserved for people seconded from the Civil Services of the member nations according to a complicated system of bargaining. I got an appointment to the Supreme Headquarters Allied Powers of Europe (SHAPE) Technical Center, as Head of the Communications Techniques branch, combined with across the board, consultancy for all the activities at SHAPE Technical Center. I was also responsible for organizing and running cross-fertilization seminars. Very soon thereafter, the Head of Communications Networks branch head became vacant, and so I was asked to do this job in addition. Shortly after that, the position head of Communications overall became vacant and I filled this job too, for nine months, until it was filled as scheduled by bargaining between the nations. As far as remuneration was concerned, I received marginally more than the Deputy Director of the Technical Center. The Director, Dr. Erik Klippenberg of Norway, wrote to me, "We feel ourselves extremely fortunate in having you at SHAPE Technical Center. You are undoubtedly enhancing our reputation and standing."
I moved to The Hague, where SHAPE Technical Center was located, immediately after retiring from GCHQ. We sold our house in Cheltenham, bought a new house, to have a home still in the UK, in Bristol where our younger son was just starting work, and moved to a hired house in The Hague, all in the three frantic days between finishing work at Cheltenham and starting work in Holland. The SHAPE Technical Center normally worked on three-year contracts, because they were keen to have people from all fifteen NATO nations – partly to draw upon their different expertise and culture, and partly to make sure that people who had learned about and got committed to NATO would go back to their respective countries, and so enhance coherence across the alliance. Of course, as in many other things, the odd man out was France, where the appellation for NATO becomes OTAN, which is NATO backwards. This is characteristic, because France was a member of NATO for policy purposes, without subordinating in any of its Forces to NATO’s military structure. However, France still insisted on being represented at all meetings, including on military matters. However, we had no French people in the SHAPE Technical Center.
In recruiting we had to give some attention to balance between nations, but our predominant objective was to get the best people we could find. There were interesting cultural differences. In recruiting young people from Europe we found they had attended longer and somewhat more formalized degree courses, and tended to be stronger on basic theory than those from the UK or USA. Even though their courses were longer, they did not normally include any project work. Age for age, their ability to appreciate the engineering environment and tradeoffs tended to be less mature than that of UK candidates with more practical experience. Therefore, among people of equal ages, the British candidates tended to be superior in doing practical work and getting solutions to practical problems. American candidates normally came from the Department of Defense civil service. Although academically quite well qualified, they were mostly professional project managers and administrators, with limited hand-on experience. We were more concerned with solving problems ourselves.
Apart from such general differences, the wide cultural diversity made for a very stimulating environment. It was perhaps a bit too stimulating, because, when people choose to interrupt their careers for a three year-attachment to NATO, it normally means they are experiencing some sort of problem at home. Most of them were very good people. Otherwise we wouldn't have them. However they were also often difficult people, prima donnas. When you have got a mixture of multinational prima donnas, life can be quite interesting. It was also interesting to observe the relation between Turks and Greeks within NATO. Individually they worked very well together, but it was not at all surprising to hear a Turk at a formal NATO policy meeting say, "I didn't quite get the hang of what my Greek colleague was proposing, but I can assure you that my government will not support it."
In the UK, I had had a lot of interaction with my military colleagues, and I had lectured the Joint Services Staff College and National College of Defense Studies in my own areas, but I had never attended Staff College myself. Now, however, I did attended the NATO Staff College at Oberammergau and found it pretty interesting.
The function of SHAPE Technical Center was to analyze problems, devise solutions an, where relevant, to demonstrate them. Another NATO body, which was then called the NATO Information and Communication System Management Agency (NICSMA), had the responsibility to procure the resultant equipment for deployment. If any project involved significant cost, it required the unanimous approval of fifteen nations before it could go ahead. That was much harder even than the approval processes in our individual countries. However, I learned after a little while that, if one could devise a solution (if necessary including a demonstration) at marginal cost, one could go ahead if one merely avoided having fifteen separate objections to it. Therefore, despite the heavy bureaucracy at the higher levels, quite a few things could be accomplished.
It was also a good opportunity to use and develop my language abilities. Learning Dutch was particularly challenging, because not only all my Dutch colleagues, but everyone one met at The Hague, spoke English, at least in subjects relevant to their own professional activities, and was rather proud to demonstrate it and keen to exercise it. At first I had to learn Dutch by the combination of a teach-yourself book, listening to other people, and borrowing Dutch translations of initially simple children's books from the public library and then progressively more ambitious books, so that I could, without too much boredom, absorb the vocabulary and grammar. I soon became quite fluent, and was able to give briefings and make presentations in English, French, German or Dutch. I acquired enough Italian for some useful interaction, but not enough for any formal presentation.
I cannot say a lot, obviously, about the specific contributions I made, but one or two examples might be interesting. NATO had quite a long time ago acquired electronic exchanges for telephone and teletype communications. With subsequent developments they were no longer a match to the then current standards and requirements. People were considering a contract involving many millions of dollars and many years of development, to try to cope with this. Even so, it was judged to be a very high-risk activity, if only because the people who had developed the original switches – that is, exchanges – were no longer available. Hence anyone who had to modify the design would have an imperfect understanding of what they were trying to modify. Therefore I approached this by using microprocessors in external stand-alone interfaces between the new type external link and the unchanged telephone exchange. These cost less than $10,000 and were developed in about six months.
However the most critical problem that faced us was to make the strategic NATO network survivable. Survivability was critical for making the deterrent work in practice and, to my mind at least, making it known that it was survivable was critical for making the deterrent work at all. This is because if someone thought they could deny NATO its command and control system, they might feel relatively immune to counteraction. I was very active in developing a viable theory, not as an academic theory, but as a real guide for design and development of survivable networks, and in developing a computer tool to monitor, manage and reconfigure a complex network, if it was impaired. I also developed operating practices that allowed the way of using the system to be adapted to its surviving capability.
The people who designed networks connected a number of nodes to each other by fixed links. Some of these were cables, some were chains of radio relays and some were satellite links. It was therefore not uncommon that, in communications between two terminals A and B, several intermediate links went through the same satellite. However, the partitioning of the task was such that no one realized that the first uplink could go directly to the final downlink, cutting out all in between. It completely ignored the fact that the intrinsic flexibility of satellites for connecting any base station within view of the satellite to any other gives a much wider ability to repair or reconfigure networks. For a very similar reason, when links were within national systems, NATO had dedicated hard-wired links between nodes. I pressed to make use of the very wide redundancy and diversity of routing options available in national networks so that, in place of this, NATO could hire a service between input and output nodes on the national carrier, subject to the national carrier demonstrating that they were able to implement this connectivity in many different ways. This therefore substituted therefore virtual nets for hard links, and so maintained service even in the face of technical or military damage. This had to be associated with providing guidance and contractual and financial incentives to the commercial or national nets to make their architecture robust and survivable.
I also developed some wider concepts. One of them related to tactical links, as distinct from the strategic links in which this survivability was so crucial. Here it is the radio path which
(a) is vulnerable to jamming or interception,
(b) may jam itself or interfere with other communications.
Therefore, I pressed hard for using cable or fiber landlines as a backbone, and merely connecting a tail link to the mobile user to the nearest node in this fixed network. Thus the length of the radio path would be enormously reduced, and its liability of being jammed or intercepted, or it interfering with other people would be similarly reduced.
The static networks themselves of cables, fibers and so on normally comprised a hierarchy of local switches linking a number of local nodes, which are interconnected in a network of area switches, which is interconnected with a network of regional switches, and then national ones. This hierarchy had been developed by the national PTTs or other service providers to make the best use of the limited capacity of copper cables. With the advent of optical fibers, which can provide far higher capacity much more cheaply, the cost of transmission went down much more rapidly than the cost of switching. Therefore I pointed out that a simpler network architecture, with fewer hierarchical levels, would lead to a cheaper overall solution, approaching what I call a “virtual ether”. This is where the system of fibers can be used to establish virtual links between any two terminals when and where required, with whatever capacity is needed at the time. This needs more sophisticated capacity allocation and switching algorithms, but it can be well supported by the wideband cables now available.
Regarding military operations themselves, there is a similar case for flexibility, where traditionally predetermined formation, operated according to predetermined plans. I felt that it is much more appropriate to respond to locally arising needs or opportunities by grouping together whatever units are available, which can jointly respond to the situation, and establishing a flexible chain of military and political command, responsive to a changing scenario. Therefore the communications must not be linked to a predetermined military hierarchy either. Most of all the wider concepts which I have mentioned are now fairly widely accepted, but twenty to twenty-five years ago it was sometimes quite hard to get people even to consider them.
Although it was then the rule that appointments were for three year, STC asked me to stay for five, the limit since I had to leave at the compulsory age of sixty-five. At that point the Norwegians, where their civil service did not have to retire until seventy, offered me the appointment as their Chief Defense Scientist. However both my wife Kathleen and I really wanted to return to the UK. Shortly before I was due to retire, I got two letters in the same week. One was from Sir Eric Ash, the Rector of Imperial College, saying that it was not widely known, but that the College was about to merge with St. Mary's Medical School and he was keen to build up electronic and similar research for medical applications. Hearing that I was due to return to the UK, he wondered whether I would be willing to become a visiting professor to help with that. The other letter was from Sir David Davies, saying it was not widely known, but he was about to become Vice-Chancellor at Loughborough University and would be leaving his job as the Head of the Department of Electrical and Electronic at University Engineering, College London, and he wondered whether, as I was due to return to the UK, I might join University College London as a visiting professor, to maintain the thrust and momentum of research activities in the area he had started. My response to both was that I would be happy to do this provided they did not object to my doing both.
I took onboard the saying that, "Old professors never die, they only lose their faculties." In Imperial College the emphasis was on medical sensing and signal processing. However, it then evolved to apply the technologies which we had developed for detecting tumors also to detecting weapons, explosives and drugs, for security purposes. After a long delay, Imperial College is now at the point of establishing a spinout company to commercialize some of these developments. At University College London I was dealing with radar, sonar, satellite earth sensing, antennas, electo-optics, microwave measurements, radar meteorology and so on. Later on I also became involved in the Open University, supervising Ph.D. work mainly in underwater acoustics. With Cranfield University, more precisely the Royal Military College of Science, I acted as external examiner for the military electronic systems engineering MSc course.
At Bristol University I initiated and supervised research projects on:
• meteor-scatter communications,
• wideband communications and in particular shifting the cost and complexity from the many mobile terminals to the base stations,
• microwave sensing for detecting buried landmines,
• microwave sensing for detecting cancer or detecting contaminants in food.
A spinout company has now been set up to exploit some of these ideas.
I am not involved in administration or in teaching at any of the universities. I merely suggest, launch, and help to direct research projects, and sometimes I co-supervise those started by other people. I enjoy helping young people to realize their potential. I have had Ph.D. students from more than twenty different countries. Unfortunately not enough from the UK. The fact that there are so many from so many backgrounds is enjoyable, but it is worrying that our own supply of future research leaders in academia or industry may be at risk.
I regularly visit the universities to discuss my own and their ideas and progress with the staff and research students involved, and then follow it up with Email notes and suggestions, and further visits. Also produce and publish novel ideas outside this framework and provide quite a significant amount of consultancy to government and industry.
I am very active on the Defense Scientific Advisory Council, the equivalent of the American Defense Science Board as a member of the Sensor Technology Board, the Communications and Information Systems Board and now the Information Superiority Board. I have also been co-opted by the Weapons Board and the Main Council, and have been chair or an active member of a large number of working groups. Indeed, I have been by far the most sought-after member to serve on these. I have also chaired a number of Ministry of Defense technical audits.
We have heard a lot about your UK career in research and beyond, your first retirement and your second retirement. Now perhaps you would tell us something about your personal and general reflections about your professional and wider life.
I have been fortunate always to enjoy what I was doing. If it were not for that, I would not have carried on into and beyond first and second retirement. What is more, successive jobs have given me new stimulation, new challenges, and new forms of intellectual satisfaction.
At ASWE during the war, results of immediate use to the fighting services were much more important than civil service procedures, which is why it was a stimulating environment in which to work. After the war, I was fortunate to play a key role in transforming the whole concept and nature of coordinating and directing the operation of multi-unit forces.
At AUWE, I had the challenge of integrating fragmented capabilities arising from the different establishments that were brought together as AUWE in the field of undersea warfare, where we in the government lab were far ahead of industry and academia in capability and understanding of the problem. In fact, we had a lot more experience and insight nationally, albeit less resources, than the Americans.
At GCHQ, we were fighting a truly challenging real-time battle of wits. We had to find ways of doing things which the opposition fondly believed to be impossible, and we had to execute them quickly. We foresaw or recognized a need, devised a way of doing the “impossible” so as to meet this need in quite a short time scale, and then saw the results. That is a very, very stimulating environment.
In the intelligence world more generally, I was the first man who really had the opportunity to integrate the technical and intellectual resources and the ideas of the three intelligence services to produce synergy, for doing things more cost effectively, or doing things that for a single service would not have been justified to do at all
In NATO I had the stimulation of working with multiple cultures and the administrative challenge of finding ways to solve operational problems and to generate cost effectiveness benefit in the face of a heavy-handed multinational bureaucracy. In academia I have the opportunity to help youngsters to broaden their horizons and realize their potential at the same time as developing my own innovative ideas. There is also the challenge of working with people who sometimes find it difficult to do the urgent in the face of the important, compared with the civil service where they always do the urgent and forget the important.
In the Defense Scientific Advisory Council I have the opportunity to combine a deep and wide knowledge of defense needs and perspectives with an independent view from outside. In the setting up of spinout companies, one can facilitate the transition from research to producing something that can be deployed, and that is of value to society as well as to the profit of industry.
I have heard it quoted: "A job in headquarters administration is like heaven: we all hope to finish up there, but not while we still have a real job to do on Earth." I personally have tried very hard, and I hope fairly successfully, to combine both. However I must admit that it was at the cost of the time and effort which I could otherwise have given to family and other outside commitments. I have always enjoyed the challenge of something being “impossible”. Almost all the projects I started, both in defense and in academia, began with someone saying something could not be done. Even in the domestic field, the suggestion that something cannot be done is a wonderful stimulant.
I have never been happy directing a group or organization without also doing some personal work. Apart from giving balance in one's activities, this is valuable because guidance or directions to others it is more acceptable when they know you to be capable of doing similar things oneself - and in fact are doing so. For this reason I have always encouraged people in the scientific hierarchy to combine guidance and direction at one level with making some personal contribution at the level below. When acting as the director or otherwise chief of an establishment, I have always tried to set aside one or two half days per week for interacting directly with people at the working level at the ‘coal face’, listening their problems and ideas and contributing to both. However, I normally let them put forward the resultant proposed solutions through the hierarchy, so as not to short-circuit the normal management line of responsibility. I also found this direct contact with people at the working level very helpful in career planning and management. I always regarded as a very important part of my function to make sure people had the range of experience and opportunities for promotion that made the best of their potential.
When chairing a committee or working part, I like to put forward suggested conclusions or decisions at a relatively early stage, so that they can be interactively refined in discussion, to converge onto a genuinely agreed version in less time. For me, one of the most fruitful times for analyzing problems and thinking of solutions is in bed at night. That is why I always keep a notepad and pen by my bed. I do not really think I am likely to forget ideas I have developed during the night, but if I do not put them down in writing, I find it hard to go back to sleep for fear that I might.
I have been fortunate in achieving a fair number of academic and other awards, including Doctor of Philosophy, Doctor of Science, Honorary Doctor of Engineering, Fellowship in the City of Guilds of London Institute (at a time when there were fewer than a hundred people who had been given this award), Fellow of the Institution of Electrical Engineers, and Fellow of the Royal Society of Arts by invitation of the Council. The term ‘arts’ is misleading. The full title is Arts Manufacture and Science. In fact in this meant arts as in artisans – i.e. practical skills in industry. I was elected Fellow of the Royal Academy of Engineering at a time when there were fewer than two hundred of us. In the Institution of Electrical Engineers I have acted as Area and Regional Chairman, member of the Electronics Division and member of Council. I was vice president of the British Acoustical Society, but I resigned just before I was due to become president because I felt, due to the change from AUWE to GCHQ, it was no longer appropriate. From the IEE or the British Institution of Electronic and Radio Engineers I received four premiums: the Marconi premium, the Heinrich Herz premium on two different occasions and the Maxwell premium. I also had the royal award of Companion of the Bath, which is the highest honor short of a knighthood. This also gave me the opportunity to attend a number of the Buckingham Palace garden parties, where my GCHQ post entitled me to be in the diplomatic enclosure. There I admired there one fellow guest: a Nigerian of magnificent physique, wearing a toga-like white robe and barefoot!
I always tried to balance my somewhat intellectual professional activities with physical activity. I have played rugby in the Portsmouth area, where half of my team were members of the Special Boat Squadron, the Navy’s equivalent of the SAS. I have a black belt in Judo and represented the County of Dorset. In so-called ground work in Judo, I came to the conclusion that victory normally goes to that contestant who is least willing to concede defeat. As a mountain climber, I led the first ascent of the Cima di Moro in the Bregaglia Alps. I still do a lot of what we call rambling and the Americans call hiking, including leading appropriate cross-country walks. I enjoy water sports – sailing, swimming, surfing, windsurfing, water skiing and paragliding – although I have not done much of the last two. Also rowing, either normal boats or Canadian or kayak canoes, and scuba diving. I am qualified as a Royal Naval Diving Officer.
Now let us go on to your more general reflections.
With all significant engineering achievements the way one organizes and manages projects is really very important. I always felt one of the most important keys to success is what I call ‘parental motivation’: that whoever leads the project regards himself as the father of the concept, and feels a personal commitment to its success. However, in many cases a person who is good technically may be less good administratively, or may feel his time is not well spent on administration. There is often a good case for a two-man partnership, where one is primarily technical and the other concentrates mainly on management aspects. It depends on personalities as well as their positions in the hierarchy as to who should be regarded as the project leader and who as his deputy. It is also important to understand the dynamically evolving technical and operational environment in which one wants to develop and deploy whatever comes from one's research or development. It is important that the scientist or engineer plays a key role in defining the objective and does not depend on specifications by the prospective user, which are often shortsighted and limited by his past experience, or by the duration in which he expects to be in his present post.
I have been fortunate to have had sea time in all classes of ships and airtime in most types of naval aircraft. Incidentally, I believe that I was the first person ever to be transferred, suspended by a rope from helicopter, to and from a nuclear submarine at sea – a rough sea at that.
It is important that, from the very start of a project, one thinks about how it is going to go through to development, manufacture, deployment and eventually in-field support. It is also important to have team continuity in seeing a project through all these phases. That need not mean that the same people will go all the way through. It may not be sensible career management for some and they may not all be suitable for it. However a significant core should carry on through consecutive phases, to maintain continuity of motivation and of understanding what was done and why, and to make sure that people engaged in the current stage also plan for the further stages in which they expect to be involved. The management processes in many government and industrial organizations can make it difficult to achieve these objectives, but it is important to aim at them.
For keeping a coherent common vision within a team, I am very keen to have small teams of high-caliber staff. This is because if you have got "n" people working together you have n(n-1)/2 interfaces. If many people have individually limited responsibilities, there are far too many interfaces, which take far too much effort just in keeping them compatible. In addition, they individually focus on a limited part of the overall concept and miss opportunities for wider tradeoffs and synergies.
Turning now from the management of individual projects to managing an overall program, I am a believer in a matrix organization, resembling a fabric held together by a warp and an orthogonal weft. The horizontal warp provides the common technological means and concepts to facilitate the synergies and tradeoffs between projects equivalent to those I spoke of before within projects, and the vertical weft links all the relevant technologies together in a single end product. Hopefully one can avoid friction between the two but, in the limit, the achievement of the end product, i.e. the weft, must take priority.
There are many political and management pressures for working to fixed specifications, schedules and budgets, but in the development of a really worthwhile step forwards significant time must elapse. In this time technologies, the end capability to be achieved, the environment in which it is to be used, scientific understanding and technological understanding of how things can be done, all are likely to change and evolve. Therefore the idea of having everything predetermined is like sending a missile on a long trajectory to a moving target with no opportunity for mid-cross guidance. In trying to optimize a development in the faced of these uncertainties, it is important to appreciate that progress takes place in a number of steps and, at each such step there are a number of possible ways one can go forward. I believe that these decisions should be guided by what I call “planned serendipity”. Thus, at each decision stage, we should pick an option which will open opportunities for evolving in many potentially relevant directions and forecloses as few as possible. In this way the chance of eventually finding oneself confined in a blind alley is minimized.
There is a lot to be said for using commercial standards, components and products, even at the expense of some adaptation of the stated requirement, so that specialized areas like the military or intelligence or others, do not needlessly duplicate work being done elsewhere. Otherwise, it is quite likely that, by the time they have developed their special technology, it may be overtaken by what has in the meantime been done, with much bigger resources, in industry. In addition commercial products enjoy continued support and further development or replacement by superior successors, which ought to be exploited.
Project management always needs monitoring how individual components, team members or contractors are performing. Whilst this is an essential management tool, it is important to recognize it is that and nothing more. In many organizations, monitoring progress becomes a self-justified bureaucratic activity, which diverts time and creative skills from making progress. Cost control of development must be balanced against cost effectiveness of the end product, and must not be an objective in its own right. Control of a single project to meet its defined objectives should not be done as a self-contained activity, in a watertight compartment, without consideration of optimizing the total program of which an individual project may be part, and of opportunities for mutual support, tradeoffs or commonalities. There is a tendency for managers to be very averse to risk. Of course they feel any failure to meet targets in performance, time or cost will reflect badly on them. However, when directing a program of multiple, reasonably independent projects, one should maximize the sum, over all projects, of the probability of success of each constituent projects, multiplied by the payoff if successful. Without taking some risk that some projects may be less than fully successful in meting their ambitious target, or may even fail altogether, the opportunities for getting really big payoffs are likely to be missed.
I have always tried hard, and often successfully, to make sure that of the order of 10 percent of research and development resources were put aside for self-initiated work, started off by researchers rather by the customers. Self-initiated work is not necessarily research. It can be development. Of course this entails the assumption that the researcher understands what his likely customer may need in the future perhaps better than the customer himself. It is normally easier for an engineer to understand the problems of the user than for a user to visualize the opportunities created by evolving technology. As stated, the 10 percent is set aside primarily for self-initiated innovative work. However, because these resources are not committed to projects on which other people already heavily depend, it also creates a y flexible reserve, for dealing with unforeseen serious problems in existing commitments and unforeseen novel opportunities as they arise. At one time when we had a new naval Captain Superintendent at ASWE, he was very disturbed to find that a significant fraction of the R&D staff was engaged in self-initiated work. He therefore got one of his Commanders to analyze all the past projects and classify them into the three classes of:
• ‘outstandingly successful’,
• ‘of doubtful value if not complete failures’.
He was somewhat surprised, as was the Commander, to find that every one of the 25 percent or so projects marked as ‘outstandingly successful’ had started with self-initiated work, and every one marked ‘of doubtful value’ or ‘failure’ had started from a formal staff requirement.
As I already mentioned, useful initiatives by the scientist or engineer require real contact with and real understanding or visualization of their operational environment and requirements, now and in the future. At one time, on a transatlantic air trip, sat next to the general manager of an American mutual fund – what we call a unit trust. He had a team of graduates from the Harvard Business School, who investigated companies in great detail before he made any decision to invest in them. However, he found that a good rule of thumb was:
“If the chief executive of a company is an accountant I can rely on good returns over a three-year period, but if the CEO ias an engineer and has the support of a good financial director, I can rely on good returns over a ten-year period.”
He added, "but of course I am only a bloody accountant."
I am sure that the IEEE and beyond will be fascinated to hear your thoughts on running an R&D establishment, especially with the vast and high-level experience you have had over the years.
I feel that ideally one wants a hierarchy in which every senior is responsible for somewhere between three and five subordinates. He himself may also handle one of these subordinate activities, provided his external burdens do not make this unfeasible. You want sufficient levels in the hierarchy to provide motivation, by making certain that almost everyone can see prospects of moving up. Too flat of a hierarchy removes this.
It is also important that people have a reasonably self-contained well-defined task, with the authority and resources for managing that task. Similarly, those lower in the hierarchy must have the responsibility, authority and resources for their own well-defined bit of the overall task. Therefore you want to give guidance to the level below without detailed day-to-day interference. The aim is to combine expertise, responsibility, authority and accountability for everyone. Ideally, all should say, with conviction, “those above me could not do my job because they do not understand the detail, and those below me could not it because they do not see the broad picture.” In every organization, there is a formal organizational chart, and there is an unofficial networking structure, by which people do their day-to-day work. I believe the criterion for a good organization is minimum discrepancy between the two. There is always a temptation to get things done internally by one's own workshop or drawing office or whatever, rather than having to find funds for placing subcontracts outside. To assure that things are done in the most effective and cost-effective manner, one must make sure that the use of internal resources is also accounted for, but this must be done by a simple rough-and-ready formula. Otherwise it creates a needless bureaucratic burden.
I always attach great importance to career planning for the staff and succession planning for posts. In many organizations, particularly the civil service, there is a tendency for automatic job rotation, where everyone is moved every three years or so to another post, for wider experience. I do not believe that either the individuals or the organization benefit from an environment where all are jacks-of-all-trades and none masters of one. The Civil Service has a limited opportunity for special-merit promotions (from which I myself have benefited repeatedly). I believe that, beyond this, one must recognize individuals who will get the most fulfillment, and give the most effective service, if allowed to develop a specialization, and to serve in that for a much longer period of time, possibly their whole career.
When I came to AUWE, a lot of regular reports were required from various parts of the organization. I abolished the obligation to provide routine reports, except in cases where there was some evidence that they resulted in some useful action that would not otherwise have taken place. I also abolished most standing committees. I found it was much more productive to replaced them, where appropriate, by limited-duration working parties with a precise objective. These working parties reported their recommendations and conclusions, and sometimes suggested an appropriate interval, after which the same subject might be revisited.. I delegated authority to sign letters with the words "For the Director" to the lowest practical level, and encouraged outside bodies, to address letters "To the Director, AUWE, for the attention of so-and-so". I put a lot of pressure on all of the staff to answer letters within 48 hours, unless there was a good reason to the contrary. Having delegated to the lowest feasible level the authority to write “on behalf of the Director”, I balanced this by requiring that all correspondence signed "For the Director" to be circulated to me and my deputies every day: not normally for detailed study, but for quality-control sampling.
There is always a balance to be struck between central and delegated allocation of resources. Taking the example of a workshop or drawing office or typing resources, if these are central, one can be sure that they are fairly fully and steadily loaded, and shared flexibly as the load varies between users, but with the disadvantage that none of the people there have a detailed understanding or a personal link with those whom they serve. If they are all spread out, one gets the opposite effect. Therefore I always try to achieve a balance, in which each group has the dedicated support facilities and staff to handle its steady ‘base load’, so as to employ this staff reasonably fully and effectively, virtually all the time. Any excess work is then farmed out to a central pool. The same principle applies to some extent to the matrix organization, where the technology strands include both experts y allocated to specific projects and others in a pool which provides common support for the same technology strand to all of the projects.
In the following section, I air some thoughts and experiences regarding the top levels of the UK Civil Service. Some of these remarks may repeat points made earlier in a more limited context. When Director at Portland, I was sent to the Civil Service Staff College. I quite enjoyed that, but learned very little that I had not already worked out for myself, and applied in my day-to-day work. I was not sure whether the course was of much benefit to those who did not have any management experience, expected to work mainly on policy advice to ministers, and might never have any management responsibilities. It did not teach the importance of having a real insight into the field which one is managing; rather, it regarded management as a skill in its own right, quite detached from the activity that is being managed. That is probably typical of the cult of the amateur, which unfortunately is very common in the UK. Planning, monitoring, reporting and other management techniques tended to be regarded as things that are desirable in their own right, rather than supporting tools for informed judgment and helpful decisions. There is a complementary relation between the urgent activities that should be handled by the chief executive or chief of staff, and the important longer-term policies and strategies, which hopefully are handled by the president of an organization or the commander in chief. However this balance was never mentioned. Most civil servants and politicians, in my experience, focus on the urgent. Most academics focus on the important. Only a minority maintains a proper balance between the two. I tried to do so, but it is not easy.
Committees are often designed to disperse or even obfuscate responsibility. Committees are useful to air issues, ratify decisions and coordinate implementation, but in most cases the useful decisions are made informally outside of the committee. Policy papers are normally and sensibly structured as an abstract, an executive summary, a main body, and a number of detailed annexes. It is important that, in writing such papers, one explicitly aims these components at successively lower levels of the hierarchy who will read and hopefully absorb the papers at the level of detail intended for them. Turning again to the Treasury, as it worked in the days when I was involved, resources were rather rigidly partitioned into single-item budgets and single-year funding releases. This militates against the benefits to the nation, in cost and performance, of an integrated program covering an entire field of related activities over the total duration of that program. In all my posts I tried my best (despite this unfavorable background) to achieve global optimization, at least over the full scope and duration of the program for which I was responsible. In fact I tried to do it over the full field, as far as I could see it, beyond my own area. In a truly closed system, the cheapest procurement of a given result is obviously the best.
The approval of projects or award of contracts often has quite important implications and side effects, not covered by the specification or included in the costs. Therefore decisions based purely on a specification and cost are likely to give non-optimal or even undesirable results. In reaching decisions, there ought to be a scope for judgment, exercised on the basis of insight. This is unpopular not only because of difficulty in predicting what decisions will be made, or of subsequent auditing, but also because this can make it more difficult to justify decisions against political challenge. Treasury personnel were all extremely able. Only those who were very bright could have got into the higher levels of the Treasury. Nevertheless, they are not often in a position to understand this sort of issue. The terms of reference of those who do understand them rarely allow them to take these wider issues into account. Almost all generic rule, and their interpretation in specific rulings, have unintended consequences. There was, at least in my time, no real effort to try to analyze the likelihood and nature of such consequences.
Ralph, you have given us your judgments on lots of organizations and how they tick. Perhaps there are some final remarks that you would like to make that have not been covered thus far.
All I would really like to say is that we all spend most of our waking hours at work. Therefore, if you want a fulfilling life, you should make your life's work also your principal hobby – not your only one, but your principal one. For me engineering, particularly at the system (or system of systems) level, has been ideal in providing such a creative and satisfying hobby.
Professor Ralph Benjamin, CB, thank you very much indeed for giving us the UKRI oral history that we can present to the IEEE in the USA. You have given us a lot of your valuable time. Thank you very much indeed.