In Search of Techne

Autobiographical Notes

(c) Roy Johnston 1999

(comments to rjtechne@iol.ie)

These notes are as I wrote them in the early 1980s, when I was preparing the book for publication by Tycooley. RJ July 2001.

The writer was lucky enough to have been able to go to a school (St Columba's College, near Dublin, 1941-46)(1) where as well as science laboratories there were workshops (electrical, electronic, wood, metal) and a farm. He considers that exposure to this range of early influences gave him his feel for technology, although he went on to take a science degree in Trinity College, Dublin (1946-51), and a mathematics degree simultaneously, which was then possible.

The science degree then started with physics and chemistry; one could specialise in either in the later years, and the writer opted for physics.

At that time in physics the tradition was that you built your own apparatus; Walton had been back from the Cavendish Laboratory in Cambridge for about a decade, which included the war. The 'string and sealing-wax' tradition of the Cavendish was alive and living in Dublin, which had managed to escape the cornucopia generated elsewhere by the needs of military technology.

After taking his degree in 1951, the writer went to France, where with the aid of a small French Government grant he had the opportunity of serving his time in 'big physics' with a team led by Bernard Gregory (who later became Director of CNRS) in the technology of high-energy nuclear physics. This, for him, was a good training in what we would now call 'systems engineering'; big science was pushing the technology to its limits in order to delve deeper and deeper into the basics of the nature of matter. The physics, while exciting (new particles were being discovered, and there were problems in characterising them) was for him of secondary interest. The art of 'getting the experiment to work' was what 'turned him on'. By nature, he found himself to be an engineer, though with the training of the physicist he was in a position to work at the frontiers of technology in the development of novel systems.

For example, the Ecole Pylytechnique cloud-chamber system was controlled by a digital electronic logical sub-system, actuated by electronic, hydraulic and pneumatic sub-systems, and involved a solenoid which dissipated half a megawatt. The industrial relevance of this type of 'big science' was by then beginning to be recognised.

Why didn't he stay there? At this point other factors intervened. There was a world conference at Bagneres de Bigorre in 1953, which brought together the 'new particle' people. It was held there because the writer's work-place, the Pic du Midi, was the location of the largest Wilson cloud-chamber in the world and one of the main centres of high-energy nuclear physics research, the source of high-energy particles being the cosmic rays, which occur in greater intensity the higher up you go.

At this conference he met with Cormac O Ceallaigh, who had just then left University College Cork to take up the Chair at the Dublin Institute of Advanced Studies. O Ceallaigh was then a celebrity, in that he had discovered and recognised what was then known as the K-meson, using a totally different technique (microscopic examination of traces left in thick photographic emulsion). What was the relationship between the K-meson of O Ceallaigh and the V-particles as seen by the cloud-chamber people? Were they the same? There were problems in measuring the mass reproducibly. O Ceallaigh wanted to set up a photographic emulsion group in Dublin and the writer agreed to join him; here was a new technological frontier to work at in the service of physics.

As it happened, he left cloud-chambers at a good time, as within a year or two they had become as obsolete as the sailing ship, with the advent of the bubble-chamber, which shifted the centre of gravity of elementary particle work to the neighborhood of the big particle- accelerators, which were then coming on the scene to rival the cosmic rays as particle source.

A word on the principles involved would not be out of place. With a cloud-chamber, you supersaturate argon with alcohol vapour, so that when a charged particle passes through, leaving ionised atoms in its track, drops of alcohol condense on the ions and you see a visible trace. The art is in the timing and the illumination and the photography, and above all making the whole thing happen when the cosmic particle arrives, a rare event, detected by sophisticated electronic triggering devices. With luck you could get a picture every few minutes; one picture in 100 or so might have a good event in it.

With the big accelerators, however, you knew the times of the arrival of the particles, since the machine produced them in pulses every few seconds, in large numbers, and with precisely defined energy. Some other detector was needed, with short recovery-time. The bubble-chamber turned up; this turns the cloud-chamber principle inside out. Bubbles form in the wake of a charged particle going through a liquid, provided it is superheated and wanting to boil, but prevented from doing so by the lack of nucleation centres. These the ions left in the trail of the charged particle provide. Bubble-chambers were made initially using propane, and eventually with liquid hydrogen, so that the nuclear interactions studied were between single particles, a clean experimental situation.

No cloud-chamber team could compete with this, so most of them went up-market to higher energies, where the big accelerators couldn't go for a decade or so; extensive air-showers and so on. The interest shifted towards the cosmology of the origins of cosmic rays; the bubble-chamber people had taken over the field for studying the properties of elementary particles.

Either way, it was 'big physics'; one had to be in Stanford or Berkeley or Culham or Geneva to be in on it.

On the other hand, with emulsion, there was the possibility of working away quietly in Dublin, with the big accelerators in Berkeley and elsewhere as particle sources, but doing the analysis locally.

This was an attractive concept, as it seemed to give a toe-hold into a scientific career in Ireland. For this, from very early, was the writer's motivation. It would be difficult to say why; perhaps a sense of wanting to 'buck the system' that had for generations been content to allow emigration of the youth and had retained the time-servers to form a conservative middle-aged population.

In the Institute of Advanced Studies the work involved high-precision optics, statistical analysis, and, latterly, on-line computing. The writer constructed the first dedicated on-line system in Ireland in the period 1959-60, at a time when there was one computer in the country, at the sugar company in Thurles. Programming was by patchboard and storage was by decatron tube; the system was based on a concept originating in Harwell. It was used to do the preliminary routine analysis of multiple coulomb scattering statistics, at a rate comparable to that with which the basic readings could be taken by an observer with a microscope, giving a gain of a factor of about 10 in speed. Thus the work in this period continued in the 'systems engineering' tradition already initiated in the Ecole Polytechnique environment.

In about the middle of this period (1953-60) an attempt was made by a Government commission to look into the feasibility of accepting a US research nuclear reactor, which was on offer in the general framework of the 'Atoms for Peace' programme of the time. A commission was set up to examine the question (on which O Ceallaigh served); it emerged that the cost to us of accepting the free gift (in terms of what we would have had to invest into the type of science to which a research reactor would relate: isotope work and so on) would have exceeded the amount spent in all other branches of science. The offer was therefore, wisely, declined. The episode left however some questions in governmental minds, perhaps, that sensitised them to the significance of the Lynch/Miller/OECD Report, mentioned in the introduction, with the result that the latter was not shelved, but led to some action, in the form of the setting up of the National Science Council a decade later.

By 1960 it looked as if there was no future in 'big science' working from an Irish base(2). It was necessary to diversify, generalise, get out into the 'real world'. No jobs for physicists as such were on offer in Ireland at the time (indeed, are there any now, outside teaching?). A public speech by the then managing director of Guinness's, extolling the virtues of applied science in industry, gave an opportunity for an approach, which eventually led to a job-offer with the Production Research Unit at Park Royal, the Guinness brewery in London. The objective of this unit, which involved an interdisciplinary team, was to attempt to modernise the production process, and where appropriate to render it continuous rather than batch. This involved various approaches to the instrumentation and control of the parameters on which the dynamics of the process depended.

The project as originally conceived could not be held to have been successful. An important limiting factor was the purity and health of the yeast, this being crucial in the Guinness quality control process, the final product (when bottled) being unpasteurised. It was established that a continuous fermentation under Guinness constraints (ie without yeast re-cycling) was dynamically unstable, and that to implement it would require a sophisticated feed-forward control system. The parameters of the latter were identified, and various modules of the necessary instrumentation and control system developed and tested. However, in view of the complexity of the projected system, the decision to revert to the traditional batch process (which is 'fail-safe', dynamically stable and robust) undoubtedly was the correct one.

Towards the end ot this period (1960-63) a chance encounter with Finbarr Donovan(3), then Sales Manager of Aer Lingus, led to an opportunity to return to Ireland. Barry Donovan had got into computing in the early days (ie mid-fifties), and there had been a tenuous collaborative link with DIAS. He was now engaged in trying to push through a deal whereby Aer Lingus would would be the European test-bed for a 'real-time reservations system' which they had developed for American Airlines, using 2nd-generation equipment. He needed to staff up with people of calibre fit to talk across to the IBM sales team, then led by Dr Paddy Doyle(4). The writer was recruited, without mainframe experience, and thrown in the deep end.

Thus began a stimulating and creative period, in which the use of the computer as an analytical instrument was pioneered, in an environment dominated by routine data-processing applications. The performance of the proposed IBM real-time reservations system (1963-4 concept) was analysed using a queue-theoretic approach in simulation mode (this in 1964 was novel; the early real-time system designers had come into real-time systems via rocket telemetry analysis etc, which involved steady data-flow. They underestimated the hostility of a stochastic environment; the prototype system developed for American Airlines saturated at one-third of its planned capacity.)

The results of the simulation influenced the design of the third-generation system eventually installed in 1967. In the meantime, the writer had switched to the economic planning department, and was able to develop computer modelling techniques appropriate to the needs of the fleet, route and manpower planning processes. A philosophy of techno-economic analysis was evolved in a 'mainframe' environment which would be recogniseable in contemporary terms under the labels 'conversational mode' and 'user-friendly': lay-recognisable models of the management decision-areas were coupled to random-accessible data-bases, for use in 'what-if' mode.

By 1970 this area of information-technology was beginning to appear applicable to a more general market, so the writer decided to go out and develop it, and also, if possible, to try to re-establish his standing within the Irish scientific community, rather than remaining in the economic planning groove. It seemed feasible to try to develop an operations research consultancy role on the fringe of the university system, and to try to steer the projects in the direction of the economic evaluation of 'new technology' situations, actual and projected.

A series of consultancy assignments were implemented on behalf of firms and State agencies, some of which were associated with student MSc projects. The core of each project was a custom-built techno-economic model of the relevant aspects of the clients' system, supplied with clients' data. The output was a set of conclusions and recommendations, based on quantitative analysis, as well as a software package susceptable of development as a management tool for use in subsequent decisions.

In all cases the analysis was 'one-off', as in that period the concept of engineered software on a dedicated micro-computer had not yet emerged. The main result of this experience was a sense of the potential importance of the university post-graduate system as a source of useful applied research. Accordingly, when the Industrial Liaison Office fell vacant in 1973, the writer took it on as a part-time consultancy contract. During the period 1973-76 a body of support was built up within the College around the 'applied research unit' concept, as outlined in the Epilogue (Chapter 6).

In between resigning from Aer Lingus (Dec 1969) and taking up the TCD Applied Research activity full-time (October 1976) the Irish Times column was a useful and interesting pot-boiler operation, generating contacts and experience which were of subsequent value in building up the Consultancy Group. Once the latter had started, however, the Column came into conflict with it, and it was necessary to give it up.

At present the writer is again independent, seeking to develop the 'Techne' concept as outlined in the Epilogue. This book, it is hoped, will help to generate the necessary interest.

NOTES

(Addenda or amendments to these notes are in italics. RJ July 2001.)

1. For some background on this institution see GK White's History. The school was founded by a group of 'improving landlords' in the 1830s to teach Irish to their sons, the better to be able to manage their estates. It was also considered useful for them to have a hands-on feel for the technologies relevant to profitable commercial agriculture, so as to be able to communicate meaningfully with their workforces. This was an enlightened philosophy, representing the thinking of an emergent pre-Famine Protestant 'national bourgeoisie', looking to the profitable development of Irish national resources.

2. But see Chapter 5.1; in fact there was a revival in the relevance of DIAS work, as a result of some technical innovations introduced by O Ceallaigh, with the result that the 'window into big science' constituted by DIAS, and underutilised by the university postgraduate system, remains open. The key development area currently is the European Space Agency; the DIAS people are engaged in developing equipment to put into a space-probe to take a close look at Halley's Comet. This took place successfully, Professor Susan McKenna-Lawlor in Maynooth also being associated; a high-tech space instrumentation firm has emerged as a result.

3. Now with Allied Irish Banks. In the 1990s he was teaching in the University of Limerick.

4. Paddy Doyle went on to help found System Dynamics ltd, Ireland's first software-house. He subsequently went independent, retaining a link with the UCD Engineering Faculty. He is currently living in retirement in Duncannon Co Wexford.

[To Irish Times Column Index] [1970s Overview]

Some navigational notes:

A highlighted number brings up a footnote or a reference. A highlighted word hotlinks to another document (chapter, appendix, table of contents, whatever). In general, if you click on the 'Back' button it will bring to to the point of departure in the document from which you came.

Copyright Dr Roy Johnston 1999