TWO CAMBRIDGE INVENTORS
by John M. Rolfe
This is an account of two men, Gordon Iles and Kenneth Craik, both associated with St John's College, and their efforts to devise and apply flight simulation to the problems that faced the Royal Air Force at the start of the Second World War.
Learning to fly has always been a demanding task and the development of the aeroplane has been accompanied by attempts to devise ground based simulations as training aids. By the start of the Second World War flight simulation had become accepted as a valuable tool for initial pilot training. The device that allowed this to take place was the Link Trainer, developed in the late 1920's by Edwin Link. Link employed technology and components used in the manufacture of player pianos, to build his first flight trainer in the basement of the family piano and organ factory in Binghamton, New York. After many vicissitudes the Link Trainer found general acceptance in both military and civil aviation. When the United States entered the War after Pearl Harbour the Link's "Blue Boxes" were being used by thirty-five air forces around the world.
While the Link served well as a device that simulated flight it did not simulate the flight of any specific aircraft. This was not a problem when training students to learn to fly basic training aircraft. However, it became more of a deficiency as the development of military aircraft accelerated in the late 1930's. The increased technical complexity of the new generations of military aircraft created new training problems. For the students these problems arose not from knowing how to fly but knowing the aircraft in which they were to fly.
After learning to fly, aircrew mastered their advanced aircraft at Operational Conversion Units (OCUs). It was here that mistakes in aircraft handling and operation resulted in the damage and loss of valuable front-line aircraft, many of which were built in North America and ferried across the Atlantic. The position is recorded in the Air Ministry publication, Notes on the History of RAF Training, 1939-44, published in 1945:
"Before the days of variable pitch propellers and retractable undercarriages, cockpit drill did not have the same significance as it did after their introduction. The trimming and handling could be explained in a quiet half hour's instruction on the ground before an instructor and pupil went into the air; the possibility of making mistakes such as taking off with the propeller in coarse pitch or landing with the undercarriage up, or raising the undercarriage instead of flaps after landing did not exist. But the position was very different after 1937, when increasing speeds of aircraft necessitated more swift and more instinctive manipulation of controls."
Cockpit drill therefore, was only fully appreciated when modern monoplanes with their extensive ancillary controls and higher speeds replaced the old biplanes. What was needed was a safe means of becoming familiar with the complexities of the new aircraft and their systems and it is in this arena that Gordon Iles made his contribution.
Gordon Butler Iles was born in 1908. He was educated at St Dunstan s Preparatory School, Burnham, St Bede's, Eastbourne and Marlborough College. He followed his elder brother Jack to St John s College in 1927 and graduated in 1933.
Gordon Iles' career at Cambridge appears not to have been marked by any singular achievements in either college or university life. However, his future was to a great extent already moulded by the time he came up to St John s. It was to be influenced by the interests of his father John Henry Iles. John Henry Iles was born in Bristol in 1871 and showed very early talent as an organist and choirmaster. He came to London with the family business but very soon changed careers to become a journalist, initially as editor of the Organist and Choirmaster . On a visit to Manchester he became attracted to the brass band movement. He bought the newspaper The British Bandsman and also began composing music for brass band. In 1900 he enlisted the support of Sir Arthur Sullivan to inaugurate the National Brass Band Championship at the Crystal Palace. He remained a supporter of the brass band movement until his death in 1951.
In 1906, while on a visit to the United States with a brass band, Iles saw a scenic railway, or roller coaster, and bought the British rights to the device. Iles built the first scenic railway in Britain at Blackpool and others followed. He went on to establish amusement parks in Barcelona, Berlin, Brussels, Cairo, Copenhagen and Paris. In 1919 he bought and developed the Dreamland site at Margate on the Kent coast and moved there with his family. For young Gordon it meant growing up in a climate of music and mechanical amusement devices.
To this mixture must be added a further ingredient. John Henry Iles had retained his journalistic interests and acted as adviser to the Aeolian Company, makers of reproducing pianos, on their publicity material. The reproducing piano is a superior derivative of the pianola or player piano in that it can reproduce not only the playing of notes and pedals but also the touch" of the live pianist.
The link with the Aeolian Company meant that the family always had a state of the art reproducing piano at home. Gordon Iles sister and brothers took little interest in the instrument, but he was frequently in trouble for taking them apart to discover their construction. In the end he was given his own reproducing piano and a workshop in which to attempt modifications. These efforts must have been of value to the Aeolian Company because when Gordon Iles came up to Cambridge they loaned him an instrument for experimental purposes. Iles continued with his developments and made regular visits to London in order to make recordings in connection with his attempts to improve the quality of reproduction of the recorded performances.
On leaving Cambridge Iles joined the Aeolian Company and became their chief theoretician and designer in the United Kingdom. He worked on the miniaturisation of the pneumatic systems used to actuate the reproducing mechanisms. However, his long term intent was to design and build an improved expression decoder.
After Cambridge Gordon Iles had developed another interest and that was flying. He learned to fly at the Cinque Ports Flying Club and joined the newly formed Civil Air Guard. So, when the Aeolian Company ran into financial difficulties in 1938 he left and joined the Royal Air Force.
For some time Iles worked as a flying instructor, but when the Operational Conversion Units began to experience aircraft handling problems of the kind mentioned earlier in this paper he was asked if he could draw upon his expertise in the player piano industry to design and build training simulators for specific aircraft types. In 1941 Iles set to work with a team of technicians from the Aeolian Company. The first simulator to be built was for the Lockheed Hudson that was in use with RAF Coastal Command. The Hudson OCU was at RAF Silloth in Cumbria and so the devices that resulted became known as Silloth Trainers.
There followed other versions of the Silloth Trainer for the Wellington, Halifax and Lancaster bombers, the Mosquito fighter-bomber and the Dakota transport aircraft. These devices were praised not only for the accuracy of their handling but also for the quality of the sound effects that were present such as engine and aircraft systems noise. The latter attribute was not surprising as Iles was able to draw upon experience gained in creating special sound effects for the repertoire of the cinema organ.
Intelligence about Iles' training devices reached the United States of America and in October 1941 plans and specifications for the Hudson Silloth Trainer were obtained from England. The U.S. Navy liked the idea and decided to build its own simulator that would duplicate, as nearly as possible, the operating characteristics of the Martin PBM Mariner flying boat. Taking its cue from the English, the U.S. Navy sent the specifications to M. P. Moller, Inc., an organ manufacturer. Development and construction began within weeks. In spite of both major and minor modifications the Moller company completed a considerably different Silloth trainer by June 1942.
For a number of reasons, both engineering and climatic, the pneumatically driven simulator did not prove reliable. Consequently the U.S. Navy turned to the Bell Telephone Laboratory to produce a flight simulator based on an electronic computer. A successful production prototype was completed by October 1942. The first operational simulator was officially accepted by the U.S. navy in March 1944. Thus indirectly Gordon Iles had a role in initiating the development of electronic computer based flight simulators, the descendants of which are the flight simulators of today.
Back in Britain, as the number of Silloth Trainers grew, it also became clear that pneumatic technology had reliability problems. The Trainers were susceptible to climatic changes, their systems had to be kept free of leaks and they required regular calibration. Gordon Iles was kept busy moving between installations carrying out maintenance and modifications. Finally, after a serious road accident, which left him unconscious in hospital for a long time, the Air Ministry insisted that Iles take less responsibility for maintaining the devices himself. He was to divide his time between designing new trainers and training the technicians who would maintain them.
The problems with the reliability of pneumatic systems finally led this country to move in the same direction as the United States and develop simulators based upon electronic computing systems. Nevertheless, some of the Silloth Trainers remained in service with the RAF after the end of WWII and continued to earn the praise of those who used them.
Kenneth Craik's contribution was to the investigation of another aspect of the problems faced by the RAF as it adapted to new aircraft at the start of WWII. Aeroplanes with higher and more sustained performance placed additional physical and intellectual demands on those who flew them. Accidents and losses not due to enemy action but to fatigue were a cause for concern to the Royal Air Force. Although the Battle of Britain was still in progress, it was evident that any future fighter operations would be of a different order. Defensive sorties over home ground would change to offensive activities, either in large scale fighter sweeps across France and the Low Countries, or as extended protective cover for bomber raids on Germany itself. This would mean that the range of the single seat fighter would need to grow and the sortie length would increase. To illustrate this point, early marks of fighter aircraft such as the Hurricane had a maximum loaded range of 525 miles while by 1942 the Mustang had a fully loaded range of 2,080 miles.
More needed to be known and understood about the effects of extending airborne operations. Consequently, the Air Ministry sought the advice of its Flying Personnel Research Committee (the FPRC). A leading member of the FPRC was Frederick Bartlett FRS, the first Professor of Experimental Psychology at Cambridge, who was an advocate of applying psychology to problems associated with working conditions.
Bartlett was concerned about the performance of military personnel and the tasks that were placed upon them. Later he drew upon experience when he wrote:
"But there are still lots of so-called practical minded people, both in and out of the fighting services, who put their faith rather in the astonishing adaptability of the human body and mind. They find a few operators who seem as if they can put up with the new strains and stresses which the new weapons impose. They assume that all healthy operators ought to be prepared to do what a few intensely interested people will force themselves to do at a cost of very great effort. Very few people seem to realise that if the added stress of real crisis should arrive all sorts of efficiencies acquired at this cost are certain to be very short lived."
(From the text of the lecture Man and the Weapons by Sir Frederick Bartlett CBE FRS to the University of London Board of Military Studies May 1955.)
On the subject of fatigue he recorded his reaction to the problem as he saw it in 1940:
"For some time I had been trying to think how the conventional laboratory procedure for the study of fatigue might be supplemented in certain ways, perhaps improved. The common methods, based upon an investigation of simple and relatively isolated muscular and mental processes, seemed to me so devised that practically only three types of result could be recorded accurately the amount of deterioration of work, checks and spurts in work, and the final collapse of work. I thought something was needed which would show clearly and exactly how skill, long continued, may change and perhaps disintegrate."
(British journal of Psychology, 35, 3, 109-116, 1946.)
To further this work Bartlett turned to one of his junior colleagues Kenneth Craik. Kenneth James William Craik was a polymath. Born in 1914 he showed great promise as a classicist while at Edinburgh Academy and subsequently went on to read philosophy at Edinburgh University. He was also an inventive and a skilled engineer. He studied psychology as a subsidiary subject and made various pieces of laboratory apparatus while an undergraduate. He also built a set of miniature steam engines, each one smaller than the last, which is now in the Royal Scottish Museum. He wrote poetry and played the violin.
In 1936, after graduating and spending a further year at Edinburgh University Craik came to Cambridge to work with Bartlett. Bartlett recalled how Craik's professor at Edinburgh had said to him Next term I am going to send you a genius." Craik completed his Ph.D in 1940 and the following year he was made a Fellow of St John s College. Craik's field of research was vision but his interests ranged widely. He revelled in applying his studies to problems presented by the armed forces. He was particularly interested in visual adaptation and problems relating to flying.
Craik was also devising and building apparatus for use in the laboratory and the field. One of these, the Control of Velocity Task (CVT), became part of the RAF s aircrew selection battery and remains so to the present, although it is now in a computer-based form. Craik was the facilitating power of the Cambridge programme. Oliver Zangwill, a contemporary of Craik and Bartlett s successor, later wrote that, without wishing to play down the role of Bartlett himself, the whole success of his war-time policy hinged on one man, and that man was Kenneth Craik.
Bartlett continued the story:
"So I asked Kenneth whether perhaps it would be possible to design an experimental cockpit, so that the essential control responses of the aircraft-pilot, flying on instruments, could be accurately recorded, if necessary for long periods, and we should know, not only whether less or more work was being done, but also by what changes in the coordinated activities these, and other variations, were brought about. He jumped to the idea. He pulled out his wonderful black wallet, stuffed to overflowing with odds and ends of bits of paper, with their jotted notes in a strange handwriting about projected experiments, with dates of engagements seemingly in a terrific muddle, with impromptu drawings of apparatus. He found an available bit of blank space. A diagram began to grow."
Craik's first idea was to use a Link Trainer. He rejected this option because the Link's pneumatic systems did not allow the opportunity to record the detailed changes in performance that took place during the experimental runs. Craik then embarked upon constructing a device of his own design. The basis for the equipment was the front end of a Spitfire that was obtained from Farnborough. Craik's approach appears to have been influenced by two factors. The first was his expertise and skill in mechanical engineering. The second was the interest in Cambridge at that time, in the application of the differential analyser. The differential analyser was a mechanical device and came in a variety of forms. The simplest was of the wheel and disc kind. A sliding carriage carried an integrating wheel that was held in contact with the face of a rotating disc. The integrating wheel was driven by the rotating disc, its speed and direction being governed by the position of the wheel about the centre of rotation of the disc. The wheel was keyed to its shaft so that its rate and direction of rotation could be transmitted to other mechanisms. By linking several wheel and disc devices in this way most complicated computations could be undertaken. Lord Kelvin used this form of integrator as the basis for a harmonic analyser 1878. In 1931 Dr Vannevar Bush built a machine for solving differential equations up to the sixth order. Between 1931 and 1939 work on mechanical differential analysers in the United Kingdom was carried out mainly at Manchester and Cambridge. Two machines were built each of eight integrators. Moreover, it was also becoming appreciated that these devices could be designed to function as simulators.
The device Craik produced was described by him as the Fatigue Apparatus but it subsequently became known as the Cambridge Cockpit. Bartlett wrote of the device:
"The whole thing was a very brilliant and beautiful application of calculating machine principles to a complex psychological problem. It was built in our own workshop, with slender resources and at trifling cost. It was to stand up to years of hard work and it opened up what may well be a new chapter in experimental psychological development. For not only did it show that 'skill fatigue ' is in many ways different from that deterioration that long spells of work may impose upon simple muscular and mental tasks, but also it demonstrated that it is possible to submit highly complex bodily and mental processes to exact and illuminating measurement."
By early 1940 the apparatus had been built and by December a report was presented to the FPRC describing the results of An Experimental Study of Mental Fatigue (G.C. Drew FPRC Report 227). The main findings of the first study have been summarised as follows:
"The tired operator tends to make more movements than the fresh one, and he does not grade them appropriately to the stimuli provided by changes in instrument readings. Furthermore he does not make the movements at the right time. He is unable to react any longer in a smooth and economical way to a complex pattern of signals, such as those provided by the blind-flying instruments. Instead his attention becomes fixed on one of the separate instruments, and he attempts to make adjustments to changes in this, leaving the others until their indications have departed far beyond the range he would normally tolerate. Although his is unaware of it, his standards of normality relax considerably, so that now a greater external change is needed to initiate a corrective response. Signals in the periphery of his field of attention tend increasingly to be neglected. Thus he will fail to respond to a falling indication of the petrol gauge by turning on to the reserve supply. He becomes increasingly distractible. And, increasingly, feelings of bodily discomfort begin to obtrude upon his consciousness. If he becomes aware of his diminished accuracy of performance, he is apt to attribute this, not to his own shortcomings, but to imperfections that have developed in the apparatus."
(From R.C. Oldfield, The Analysis of Human Skill. In Halmos & Iliffe (eds) Readings in General Psychology, Routledege & Kegan Paul 1959).
From then on a very extensive range of studies was undertaken. The variables examined included, besides prolonged working periods, noise, sleep deprivation, the administration of alcohol, amphetamine, and vitamins. A detailed description of the experimental programme was published by HMSO in 1948 under the title "Pilot Error: Some Laboratory Experiments." Iles and Craik were pursuing their individual objectives - one to develop simulators for advanced training, the other to create effective simulated flight environments in which to undertake research. Nevertheless, their paths were converging. By the end of 1940 Craik was looking for means of building a more advanced cockpit unit in which to conduct further studies. In 1941 Craik and Iles were exchanging information about the design of their respective devices and Craik visited RAF Silloth. Subsequently, Craik concluded that a Silloth Trainer, configured as a Spitfire, would do the job and he asked the Air Ministry to provide one at Cambridge. However, nothing came of this line of action.
Craik's visits to Iles did result in a series of studies, along the lines of the Cambridge Cockpit work, which were carried out in a Silloth Trainer simulating a Wellington bomber. In these experiments crews flew a five hour sortie on instruments simulating a flight to Germany from a station in Norfolk. The results of these experiments are also reported in Pilot Error.
In retrospect, in neither Iles' or Craik's case was their involvement with flight simulation anything other than an episode in an otherwise crowded life. For both men achievements in other areas have established their lasting reputations.
Gordon Iles left the RAF after the War with the rank of Wing Commander. For a time he joined the J Arthur Rank Organisation and worked on special effects for the film studios. But the reproducing piano was still Iles first love. The Aeolian Company's only postwar interest in this area was in the manufacture of music rolls through its subsidiary the Universal Music Company. Iles acquired this part of the business and after various negotiations and mergers he founded the Artona Music Roll Company in Ramsgate Kent. It was a very small business and he once remarked that he could not afford to hire an assistant, since such would have to be a pneumatic and musical genius content to survive on a diet of bread and marmalade. The Artona company continued to produce high quality music rolls while Iles carried on with his development of an improved expression decoder. In 1973 a device called the "Robot" was completed which was free-standing and could be attached to the best grand pianos to allow uninhibited concert hall reproduction of historic performances by the original artists. The expression decoder of the Robot was designed and built by Gordon Iles. The Artona company continued to produce music rolls until 1982 when advancing age forced Iles to dispose of the assets of the company. He suffered a series of strokes shortly after and died in August 1983.
Anyone interested in the quality of performance produced by the reproducing piano may like to know that a series of compact discs recorded with Iles expression decoder equipment were due to be issued on the Nimbus label in 1996. Gerald Stonehill who has been involved in the production of these recordings, and who is a leading member of the Player Piano Society, has assessed Gordon Iles achievement as being that he put the brain into the player piano.
Kenneth Craik's reputation as an outstanding experimental psychologist grew. In 1944 the Medical Research Council established its Applied Psychology Unit at Cambridge and Craik was appointed the first Director. It looked certain that for Craik a postwar Britain held academic recognition and advancement. It was not to be. On May 7th 1945, the eve of VE Day, Craik was fatally injured in a road accident in King's Parade Cambridge. He was 31 years of age. His colleague Oliver Zangwill, who succeeded Sir Frederick Bartlett in the Chair of Experimental Psychology at Cambridge, later wrote:
"Kenneth Craik was the kind of man who, in his life and work, embodied what John Henry Newman so aptly called the Idea of the University."
The work undertaken by Gordon Iles after leaving Cambridge and the achievements of Kenneth Craik at Cambridge, deserves to be remembered as part of the history of flight simulation and its applications. Moreover, it is a testimony to the ingenuity and application of two very remarkable individuals who used their many skills to undertake applied research to benefit those who served this country in time of war.
Some ten years ago I met Professor Richard Gregory FRS at the Royal Aeronautical Society and mentioned to him that I was, then, working on a paper on the RAF's use of flight simulation during WWII. Knowing that he was at Cambridge during some of the time when the Cockpit was in use I asked for any recollection he might have of it. His reply was totally unexpected. He told me that what remained of the Cambridge Cockpit was at Bristol University. Richard arranged for me to see the Cockpit, which was stored in the basement of a house owned by the University. The Cockpit was in a very bad way. The photographs I took of it showed that it had suffered from the attentions of time and component hungry research students. Nevertheless, it was clearly based upon part of a Spitfire and retained some of the instrumentation along with a few pieces of the operating mechanisms.
Richard Gregory and I attempted to generate interest in preserving the remains, and possibly attempting a limited restoration. We had no success and shortly afterwards the Cockpit disappeared and we concluded that it had finally been broken up.
I have received much willing and valuable assistance in preparing this article and record my grateful thanks to:
Liz Ingle and Robert Fishwick of the Department of Experimental Psychology at Cambridge for access to the Craik files.
Lady Jeffreys (Bertha Swirles) for personal recollections of Kenneth Craik and the differential analyser.
Dr Alison Pearn of the Biographical Office at St John's College for information on both Kenneth Craik and Gordon Iles.
Gerald Stonehill for information on Gordon Iles' contribution to the development of the player piano.
Penny Ward, Heritage Officer at Margate Library, for information on the Iles family.