edit

Main > Archive > Early History

A Brief History of the Beginning of Operations Research in South Africa

Compiled by G. Geldenhuys with the co-operation of G.J. Rudolph and with contributions by J.W. Grobbelaar and I. du Plooy

International Background
OR in South Africa from the Fifties until approximately 1969
The Founding Meeting of ORSSA
The First Branches
The First Year of ORSSA
The Tom Rozwadowski Medal
Acknowledgements
Endnote
References


International Background

The origins of Operations Research (OR) during the Second World War (1939 - 1945) to support the Allies' war efforts have been documented extensively, both with regard to the United Kingdom (for example [27; 36] ) and the United States of America (for example [29; 30] ). In the United Kingdom Prof. P.M.S. Blackett provided strong leadership in a group which was sometimes facetiously referred to as "Blackett's circus". It is of interest to note that some South African scientists were members of "Blackett's circus", for example Professors Solly Zuckerman [36] and F.R.N. Nabarro [28]. At the outbreak of the war another South African, Dr. B.F.J. Schonland, professor in geophysics at the University of the Witwatersrand and an internationally recognised expert on lightning, was responsible for organising the Special Signals Service for the South African Defence Force, especially with regard to the development and application of radar. He was later seconded to the British armed services with the rank of brigadier and became superintendent of the British Army Operational Research Group. Later he became scientific adviser to General B.L. Montgomery with the 21st Army Group during the invasion of Europe in 1944 [26].

The first learned societies for OR were established after the war. The first was the Operational Research Club (later the Operational Research Society) in 1948 in the United Kingdom. In the United States of America the Operations Research Society of America (ORSA) followed in 1952 and the Institute for Management Sciences in 1954. The respective main journals of these three societies, namely the Operational Research Quarterly (later renamed as Journal of Operational Research), Journal of the Operations Research Society of America (later renamed as Operations Research) and Management Science, appeared roughly at the same time. Together with journals such as the Naval Research Logistics Quarterly (later renamed as Naval Research Logistics) they helped to introduce the theory and applications of OR to an international audience. With the increasing interest in OR there also arose a need for more international co-operation. The first international conference on OR was held in England in 1957 [12] and this was followed in 1959 by the founding of the International Federation of Operational Research Societies (IFORS) [30]. At the first of these conferences there was only one attendee from South Africa (and in fact from Africa), namely R.R. Tusenius [12; 32], then of the CSIR and later professor of Business Management and Administration at the University of Stellenbosch. The most important publication of IFORS in that period was probably International Abstracts in Operations Research, which from 1961 published abstracts of articles and books on OR from all over the world. Abstracts of articles and books which appeared up to 1961 may be found in the four volumes of J.H. Batchelor's well-known bibliographies [9].

(Back to the Top)


OR in South Africa from the Fifties until approximately 1969

The South Africa Prime Minister during the Second World War, General J.C. Smuts, was held in high esteem by scientists because of his personal contributions to science and philosophy. He was a Fellow of the British Royal College of Science, and in fact was its President during their centenary year, 1931. He was also a member of the British War Cabinet [10]. He recalled Dr. Schonland from the European theatre of war and appointed him as Scientific Adviser to the Prime Minister in January 1945. He was assigned the task of formulating plans for developing the country's natural resources to the full, and to co-ordinate scientific research in the national interest. One of the results of this task was the establishment of the Council for Scientific and Industrial Research (CSIR) later that year, with Dr. Schonland as its first President (1945 - 1950) [26]. Operations Research did not figure as a separate entity in the various institutes and laboratories of the CSIR which were established initially. Nevertheless, under the leadership of Dr. Schonland the multi-disciplinary approach that was a prominent feature of the early days of OR, also became the established practice in much of the applied research that was conducted by the CSIR.

In the first years of OR in South Africa there was a strong emphasis on the applications of statistical methods [1; 34], and it was sometimes difficult to distinguish between OR and applied statistics. Because of the importance of the mining industry in South Africa, it is not surprising that many of the applications of OR/Statistics were in this area. Pioneering work in this regard was done by Drs. D.S. Krige en H.S. Sichel [13; 35]. Dr. Sichel established the Operational Research Bureau in 1952, a consulting company which for nearly 20 years applied OR in a wide range of fields, including health hazards in mining, queueing problems in open-cast mines, market research, sampling and ore evaluation in underground mining, and preventive maintenance in industry. He may truly be called the father of OR in South Africa [14; 15]. Since at least the nineteen-fifties the South African Defence Force also had an Operational Research Team which was largely concerned with equipment and problems of strategy and tactics [34]. Another early success story of OR in South Africa relates to the PUTCO bus company in Johannesburg. The number of accidents on buses was reduced by introducing a selection scheme in choosing drivers, based on the minimisation of the number of accidents. An interesting feature emerging from this study was the finding that the interval between accidents, rather than the number of accidents, was the fundamentally significant variable. Iscor also applied OR since the early fifties, mainly in the operation of its open-hearth furnaces and steel mills. The South African Railways was another organisation which during this period saw the advantages of OR with regard to routing, the ratio of traction power to rolling stock, the right number of tracks on a given route, and general forward planning [1].

In the fifties the first textbooks on OR were published (for example [11; 31] ). The availability of books such as these in South Africa was a stimulus for OR.

Gerhard Geldenhuys, a lecturer at the University of Stellenbosch, introduced OR in lectures for third year Applied Mathematics students at his university in 1961.

At the CSIR in Pretoria the mathematical sciences (in the Divisions for Mathematics under Dr. A.P. Burger and Statistics under Dr. H.S. Steyn) were accommodated until 1960 under the National Physical Laboratory (NPL). In 1961 the National Research Institute for Mathematical Sciences (NRIMS) was established with Dr. Alewyn P. Burger as its Director, a position which he held until 1973 [26]. Dr. Burger was internationally recognised for his work in meteorology. One of the first major tasks of NRIMS was a large project for the South African Maize Board on the distribution of maize in the country, which could be described as a transportation problem in linear programming with the complication that the model was also expected to make pronouncements on the economic sustainability of mills. In this period Dr. Burger also contracted Geldenhuys for two projects in NRIMS, namely to conduct a survey of the possible applications of OR in South Africa, and to present a series of lectures on OR for the personnel of NRIMS. Geldenhuys used a part of his sabbatical from the University of Stellenbosch, from November 1962 until March 1963, to complete these two projects. His report on the possible applications of OR [17] covered numerous areas, many of which had received no attention in South Africa at that time. (By the year 2000 literally all the areas in the report had received attention from Operations Researchers in South Africa, whilst various techniques and application areas had been developed which were totally unknown in 1963.) Geldenhuys' lectures were published as an internal NRIMS report [16] and they were attended by Dr. Burger himself and by personnel such as Dr. Gerhard J. Rudolph, Dr. C.G. (Cas) Troskie, Dr. N.F. (Nico) Laubscher, Gideon de V. de Kock and A.P. Louw Kotze, as well as by a bursary holder of CSIR, J.W. (Jos) Grobbelaar, at that time an M.Sc. student at the University of the Orange Free State doing vacation work at NRIMS. For most of these personnel this was their first exposure to and training in some of the techniques of OR.

On the basis of Geldenhuys' report on the possible applications of OR, Dr. Burger (with the support of Laubscher, Rudolph and Troskie) decided to establish OR as a discipline within NRIMS and to appoint personnel to work in this field. Later a separate Division for Operations Research and Statistics was established within NRIMS. In 1964 Dr. Burger sent some of the personnel overseas for postgraduate training in OR. This included Gerhard Rudolph (England and the USA), Jos Grobbelaar and Louw Kotze (England).

The role of the CSIR in the early days of OR in South Africa, and the positive effect of the support that Dr. A.P. Burger provided for OR, can hardly be overestimated. On the one hand NRIMS was at the cutting edge of interesting applications of OR. On the other hand members of NRIMS were in due course appointed to positions at various South African universities and industries, where they made major contributions towards establishing OR as an academic discipline and a valuable field of application.

After his sabbatical, Gerhard Geldenhuys returned to the University of Stellenbosch, where he introduced OR in graduate and post-graduate courses in Applied Mathematics. In the period 1965-1968 he also supervised five M.Sc. theses on topics in OR such as search theory, traffic flow problems, critical path scheduling, quadratic programming (H.C. de Kock) and game theory. Cas Troskie was appointed at the University of Cape Town as senior lecturer in Statistics within the department of Mathematics, and in 1968 as its first professor of Statistics. He immediately gave attention to techniques of OR. Gideon de Kock was appointed as lecturer in Mathematics at the University of Stellenbosch and became involved in interesting applications of linear programming [18]. Louw Kotze accepted a position with Iscor where he could apply techniques of OR in the steel industry. Gerhard Rudolph was appointed as senior lecturer in the Department of Statistics at the University of South Africa in April 1967, where he was responsible for OR. He took the initiative to introduce the first degree course in OR at a South African university in 1968, at post-graduate level [21]. Already in 1968 there was an astounding number of 68 students who enrolled for the course. Other appointments at Unisa in support of Rudolph were Jos Grobbelaar and Gideon de Kock (both in 1969). Hendrik I.D. du Plessis, with experience at the CSIR in the NPL and the Division for Industrial Economics, was appointed in Unisa's Graduate School of Business Leadership. Gerhard Rudolph was appointed as professor of Statistics at the University of Rhodes in 1969. Mr. Louis Rive, the Deputy Postmaster-General in that time, knew of the possible benefits of OR because he had been Secretary of Defence shortly before. In the course of 1969 he recruited Prof. Rudolph to form an OR group in the Department of Posts and Telegraphs. Dr. Rudolph took up this task in 1970. Early in the seventies Hendrik du Plessis was appointed at Sanlam in Bellville, where he established a division for OR.

At universities other than Unisa, OR at the end of the sixties was mostly part of other degree courses, mainly in Graduate Business Schools and departments of Industrial Engineering. Shortly afterwards some universities followed the lead of Unisa and introduced postgraduate degrees in OR, for example at the Universities of Cape Town, the Witwatersrand and Stellenbosch. In 1971 Dr. H.S. Sichel was appointed as professor in Statistics and Operations Research in the Graduate Business School of the University of the Witwatersrand [14].

Another important development in the sixties was that a number of young operations researchers returned from studies overseas where they had training in OR at distinguished universities, to the benefit of their respective fields of employment in South Africa. Amongst them were D.D. (Dave) Masterson in the banking sector, H.J. (Hennie) Venter at the University of Potchefstroom, R.J. (Roelf) van den Heever at the University of Pretoria, J. Dewald Roode at the Atomic Energy Board, Gideon J.J. van Zyl at the University of Stellenbosch and later at the University of Port Elizabeth, and N.J. (Nic) Breytenbach at the University of South Africa. Prof. Venter was the supervisor for several Ph.D. students in OR topics, amongst others J.M. (Giel) Hattingh in quadratic programming and M.J. (Maarten) Venter in inventory control.

Another important development was the formation of OR groups. By 1968 a Statistics and Operations Research Group was already active at the University of Cape Town. In Johannesburg there was the Johannesburg Operations Research Group. In the late sixties these groups met regularly, usually with speakers who discussed interesting aspects of OR. The Johannesburg OR Group was constituted formally on 26 September 1968. The following office bearers were elected: RT Rozwadowski (Chairman), Dr JA Ryder (Vice Chairman), MFC King (Honorary Secretary) and D Masterson (Honorary Treasurer).

In 1968 Prof. George Hadley of the USA visited South Africa. At that time he was recognised internationally for his excellent textbooks on linear programming [23], inventory control [25] and non-linear and dynamic programming [24]. These books were well known and widely used in South Africa. He gave lectures on these topics in Cape Town, Pretoria and Johannesburg which were attended by several operations researchers. These lectures by a respected expert in several fields of OR were an inspiration for the people who attended them.

By the end of the sixties the CSIR, Iscor, the Railways, the Operational Research Bureau and the South African Defence Force were actively involved in OR in South Africa. By that time there were also several organisations doing projects which would now be regarded as OR, but which at that time was not necessarily labelled as OR. These projects were either done in house, usually by statisticians and sometimes by engineers, or externally by consultants. Amongst these organisations were AECI, the Chamber of Mines and several mining companies, oil companies (for example Shell and Caltex), Sasol, banks and computer companies.

On 18 April 1968 a meeting of individuals interested in OR was held at the Gatehouse at the Sunnyside Hotel in Johannesburg, as a result of initiatives taken by Dave Masterson, Jonathan Miller, Alan H. Munro and John C. Joslin with the support of Dr. H.S. Sichel. The attendance at this multi-disciplinary meeting was in excess of 180. The guest speaker was Prof. B.H. Patrick Rivett of England, an excellent speaker and at that time one of the best known operations researchers in the world. (During his visit Rivett also presented seminars on OR in Cape Town, Durban and Johannesburg. His father was a minister of religion who worked for some time in the Zuid-Afrikaansche Republiek (Transvaal), where he also met President Paul Kruger.) Some of the other speakers were Prof. F.R.N. Nabarro, at that time the head of the Department of Physics at the University of the Witwatersrand, and Prof. J.E. Kerrich, head of the department of Statistics at the same university. Out of this meeting a National Steering Committee was formed which would be responsible for the organisation of the possible establishment of an Operations Research Society in South Africa [28]. The committee consisted of Dr. H.S. Sichel (Chairman), Profs. G.J. Rudolph of Rhodes University, H.J. Venter of the University of Potchefstroom and C. Jacobs of the CSIR, and Messrs. H.I.D. du Plessis and J.W. Grobbelaar of Unisa, and Peter C. Pirow, R.T. Rozwadowski and J.C. Joslin of Johannesburg [33]. The group worked very hard and contacted individuals throughout the country to obtain as large a representation as possible at the founding meeting. Eventually it was decided that the meeting would be held by the end of November 1969 at the Dorothy Susskind Auditorium of the University of the Witwatersrand.

(Back to the Top)


The Founding Meeting of ORSSA

The Operations Research Society of South Africa (ORSSA) [Afrikaans: Die Operasionele Navorsingsvereniging van Suid-Afrika (ONSA)] was founded in Johannesburg on Friday, 21 November 1969. About 150 individuals were present, coming from all parts of South Africa as well as Zambia and Lesotho. As could be expected of a new society, there was a lively and comprehensive discussion of the proposed constitution, with more than twenty amendments being proposed. The constitution was eventually adopted, and the following office bearers were elected in terms of the constitution: Dr. H.S. Sichel (President), Mr. J.W. Grobbelaar (Vice-President), Mr. J.C. Joslin (Secretary), Mr. D.D. Masterson (Treasurer), Mr. M.C.F. (Mike) King (Editor), and Profs. G.J. Rudolph and C.G. Troskie (Additional Members) [22; 33]. The new society was privileged to get an excellent management team under the leadership of Dr. Sichel, who was known internationally for his work in mining statistics.

At the conference ten papers were presented, all of very high quality. A noteworthy characteristic of the conference was the excellent blend that was achieved from the industrial, commercial and academic spheres, both in the papers that were presented and in the individuals who attended the conference. The conference fees, which also covered copies of the papers and lunches, were R3,50 per person [19; 20].

(Back to the Top)


The First Branches

The Johannesburg Operations Research Group went over seamlessly into the first branch of ORSSA. The following executive committee for the Johannesburg Chapter was elected on 18 February 1970: D.D. Masterson (Chairman), J. Buttery (Vice-Chairman), dr. U. Broccardo (Secretary), M. Roberts (Treasurer), Dr. J. Ryder, A. Fletcher and C. Sankey (Additional Members) [3]. A new Pretoria Chapter was founded on 19 March 1970, with the following executive committee: Dr. J.D. Roode (Chairman), Dr. R.J. van den Heever (Vice-Chairman), G. de V. de Kock (Secretary), A.P.L. Kotze (Treasurer) and A. Haller (Additional Member) [4]. The Western Cape Branch was founded on 14 August 1972, with the following executive committee: Volker Rademacher (Chairman), H.I.D. du Plessis (Vice-Chairman), G. Geldenhuys (Secretary), J.C. Lawton (Treasurer), J. Bennett, M. Green and C. Smith (Additional Members) [7]. Branch chairmen also served on the national Executive Committee of ORSSA.

(Back to the Top)


The First Year of ORSSA

One of the main tasks of the Executive Committee in its first year was to decide on membership issues. By 16 June 1970 membership had been granted to 75 individuals, of which 47 were for full membership and 28 for associate membership. The new members included individuals who had served in the Steering Committee or who would later serve in the Executive Committees of ORSSA or its branches. Amongst them were U. Bruccardo, G. de V. de Kock, G. Geldenhuys, J.W. Grobbelaar, A. Haller, J.E. Kerrich, A.P.L. Kotze, J.C. Lawton, D.D. Masterson, J. Miller, A.H. Money, A.H. Munro, L.J. Raubenheimer, J.D. Roode, R.T. Rozwadowski, G.J. Rudolph, H.S. Sichel, R.J. van den Heever, G.J.J. van Zyl, J.H. Venter, M.J. Venter (full members) and K.C.W. Sandrock and J.S. Wolvaardt (associate members) [2]. (Two other individuals who would later play prominent roles in ORSSA and receive international recognition for their work in OR, were awarded membership in the early days of ORSSA, namely L.P. Fatti (1971) [5] and T.J. Stewart (1972) [6].

The second annual congress of ORSSA was held in November 1970 at Iscor in Pretoria, where Jos Grobbelaar was elected as the new President [22]. The other members of the new Executive Committee were D.D. Masterson (Vice-President), Dr. H.S. Sichel (Past President), Dr. J.D. Roode (Secretary), Dr. G.J. Rudolph (Treasurer), M.C.F. King (Editor), and J. Buttery and J.C. Joslin (Additional Members). In the new term of office two new branch chairmen also served in the Executive Committee, namely Dr. D.M. Hawkins (Johannesburg) and Dr. R.J. van den Heever (Pretoria).

(Back to the Top)


The Tom Rozwadowski Medal

Tom Rozwadowski was one of the main driving forces behind the founding meeting of ORSSA, and he also made major contributions at the founding meeting itself. He worked for Control Data Corporation (CDC). In 1970 CDC transferred him to their head office in Minneapolis to enable him to work towards a Ph.D. in Computer Science. He and his wife and children went on a weekend outing to Annandale, near Minneapolis, where he was to address a meeting of the Roman Catholic Church on South Africa. Due to a tragic gas leak, the entire family died at their stopover place. At the Annual General Meeting of ORSSA in 1970 it was announced that in memory of him an annual Tom Rozwadowski Medal would be made available for a paper of outstanding merit delivered by a member of ORSSA [8]. This coveted and prestigious medal of ORSSA was first awarded in 1971, to Dr. Gerhard Rudolph [22].

(Back to the Top)


Acknowledgements

The compiler of this brief history, Gerhard Geldenhuys, expresses his sincere thanks to Dr. Gerhard Rudolph, a former President of ORSSA, for important contributions which he made with regard to the roles of the CSIR and Unisa and the work of the Steering Committee of 1968/1969, as well as for his thorough and valuable comments on a draft version of the history. Furthermore, Prof. Jos Grobbelaar, the second President of ORSSA, and Mrs. Ilze du Plooy, who was the Secretary of ORSSA for many years, are thanked for their valuable contributions. However, mistakes and shortcomings in the history remain the responsibility of the compiler.

(Back to the Top)


Endnote

There is an error in the original minutes of the founding meeting [33]. According to these minutes, ORSSA was founded on Thursday, 20 November 1969. It was, in fact, on Friday, 21 November 1969.


References

1. Anderson, J.C. 1958. "O.R.." Industrial Review of Africa, October, 17-19, 21-22.

2. Anonymous. 1970. ORSSA/ONSA Bulletin, July/Julie.

3. --------------. 1970. ORSSA/ONSA Bulletin, March/Maart.

4. --------------. 1970. ORSSA/ONSA Bulletin, April.

5. --------------. 1971. ORSSA/ONSA Bulletin, May/Mei.

6. --------------. 1972. ORSSA/ONSA Newsletter/Nuusbrief, May/Mei.

7. --------------. 1972. ORSSA/ONSA Newsletter/Nuusbrief, September.

8. --------------. 1972. ORSSA/ONSA Newsletter/Nuusbrief, October/Oktober.

9. Batchelor, J.H. 1959 -1964. Operations Research: An Annotated Bibliography, Volumes 1 (1959), 2 (1962), 3 (1963), 4 (1964). St. Louis University Press.

10. Cameron, T. 1994. Jan Smuts: An Illustrated Biography. Human & Rousseau.

11. Churchman, C.W., R.L. Ackoff, and E.L. Arnoff. 1957. Introduction to Operations Research. John Wiley.

12. Davies, M., R.T. Eddison, and T. Page (editors). 1957. Proceedings of the First International Conference on Operational Research. The English Universities Press.

13. Fatti, L.P. 1988. "Current practice of operational research/management science in South Africa." Omega 16: 181-187.

14. Fatti, Paul. 1990. "Profile: Herbert Sichel." Newsletter of ORSSA/Nuusbrief van ONSA, May/Mei, 9-10.

15. ------. 1995. "In Memoriam: Herbert Sichel, 1915 - 1995." Newsletter of ORSSA/Nuusbrief van ONSA, December/Desember, 43-45.

16. Geldenhuys, G. 1963. Notes on some important techniques of operations research. Pretoria: National Research Institute for Mathematical Sciences, CSIR.

17. ------. 1963. Toepassings van ondernemingsanalise. Pretoria: Nasionale Navorsingsinstituut vir Wiskundige Wetenskappe.

18. ------. 1992. "Modelbou in operasionele navorsing." Pythagoras 30: 16-18.

19. ------. 1994. "Die stigtingsvergadering van ONSA." ORSSA/ONSA Newsletter/Nuusbrief, July/Julie, 14-15.

20. ------. 1997. "Twee konferensies." Newsletter of ORSSA/Nuusbrief van ONSA, October/Oktober.

21. Grobbelaar, J.W. 1972. "Kursusse in operasionele navorsing by Unisa." ORSSA/ONSA Newsletter/Nuusbrief, April, 2-4.

22. Grobbelaar, Jos. 1990. "Nog herinneringe uit die ou dae." Newsletter of ORSSA/Nuusbrief van ONSA, July/Julie, 10-12.

23. Hadley, G. 1962. Linear Programming. Addison-Wesley.

24. ------. 1964. Nonlinear and Dynamic Programming. Addison-Wesley.

25. ------, and T.M. Whitin. 1963. Analysis of Inventory Systems. Prentice-Hall.

26. Kingwill, D.G. 1990. The CSIR: the first 40 years. Pretoria: CSIR.

27. Kirby, M.W., and R. Capey. 1998. "The origins and diffusion of operational research in the UK." Journal of the Operational Research Society 49: 307-326.

28. Masterson, Dave. 1990. "Some recollections of the early days." Newsletter of ORSSA/Nuusbrief van ONSA, May/Mei, 17-18.

29. McArthur, C.W. 1990. Operations Analysis in the U.S. Army Eighth Air Force in World War II. Providence, Rhode Island: American Mathematical Society.

30. Morse, P.M. 1977. In at the Beginnings: A Physicist's Life. MIT Press.

31. ------, and G.E. Kimball. 1951. Methods of Operations Research. John Wiley.

32. Page, T. 1957. "First International Conference on Operations Research, Oxford, England, September 2-5, 1957." Operations Research 5: 863-871.

33. Pirow, P.C. 1969. Notule van vergadering gehou in die Dorothy Susskind Gehoorsaal om 4:00 nm. op Donderdag 20 November 1969. Johannesburg.

34. Sichel, H.S. 1957. "Operational research in South Africa." Bulletin of the International Statistical Institute 35: 391-392, 394.

35. Ventura, E.M. 1964. "Operations research in the mining industry." In Progress in Operations Research, Volume II, ed. D.B. Hertz and R.T. Eddison. New York: John Wiley.

36. Waddington, C.H. 1973. O.R. in World War 2: Operational Research against the U-boat. London: Elek Science.

(Back to the Top)

edit
Page last modified on April 24, 2018, at 11:12 AM