Um tölvuvinnslu í þágu vísinda og tækni
Minningarbrot Helga Sigvaldasonar (PDF skjal) Helgi Sigvaldason. Október 2012
Upphaf tölvuvæðingar hjá Náttúrufræðistofnun Íslands (PDF 1,7MB) Ágúst Úlfar Sigurðsson. Nóvember 2007
Fyrstu ár Reiknistofnunar Háskóla Íslands RHÍ (PDF 2MB) Magnús Magnússon. Janúar 2007.
Síldin. Upphaf tölvuvinnslu gagna Hafrannsóknastofnunar Íslands (PDF 19K). Jakob Jakobsson og Gunnar Stefánsson.
Samantekt á ensku vegna ráðstefnu um sögu tölvunnar á Norðurlöndum 2003:
Scientific computing and software development methods (Oddur Benediktsson)
Some examples of projects undertaken using the IBM 1620 computer (Magnús Magnússon)
Landsvirkjun (Egill B. Hreinsson)
The scientific and engineering community in Iceland started to use the first computer at the University of Iceland in 1964, as soon as it was installed. Some scientists had already learned to use computers at universities in other countries. A steady stream of courses in FORTRAN II was offered locally. In a year or two a number of applications had been written or acquired from the IBM Program Library or other sources. From the start research was conducted on software development methodologies (Benediktsson 1967, 1974, 1977.) Meteorologists, astronomers, geophysicists, fish- and livestock researchers, medical researchers, statisticians, and many others brought in their programs and data to be run at the University Computer Centre and to seek expert assistance. On the engineering side the use of the computer proved particularly advantageous for surveyors and naval architects with their needs for massive computations. A number of operational research models were built and simulated such as to optimise the control for water reservoirs for hydroelectric power stations and to find the optimum size of a trawler in Icelandic waters.
I shall now mention some of the many and varied projects initiated using the computer. Some of them were quite successful, others not.
In addition to holding Fortran programming courses, emphasis was put on encouraging and assisting people in the use of the computer in their own work.
At the University the first project was to computerise the calculation of the Almanac for Iceland, which had been calculated by hand and by the use of electric calculators. This enabled a great improvement to be made in the form and scope of the Almanac. Data from the Magnetic Observatory in Reykjavik were processed and analysed on the computer. Other applications followed.
Engineers were introduced to program packages such as COGO (Coordinate Geometry), a civil engineering software package, and CPM (Critical Path Method) or PERT (Program Evaluation and Review Technique). The staff of the City Engineering Department used COGO extensively in their surveying and planning work. CPM/PERT was not much used. The business administration software package COBOL (Common Business Oriented Language) did not catch on, due to lack of interest by business administration people.
The Public Roads Administration and the staff of the Computing Centre worked out the layout of the first major road in Iceland, between Reykjavík and Keflavík Airport, using the 1620.
The staff of the State Electricity Authority began using the computer in their geodetic survey work. A major long-term project on the simulation of electrical power systems was undertaken in cooperation with the National Power Company. This involved the optimisation of the use of reservoirs and thermal resources. (Sigvaldason 1967)
Several engineering firms started using the computer. This revolutionised their work in many ways.
A research project in human genetics was initiated in 1965 at the Computing Centre, supported by the US Atomic Energy Commission. This involved an extensive record linkage programme, at least by the standards of the time, covering persons in the Icelandic National Census of 1910 and those borne since. The census records were linked, using the Symbolic Programming System (SPS), to the National Register of living persons or to death records, recording inter alia the cause of death. Similarly the birth records from 1910 onwards were linked to the National Register or to death records. Thus a database of the 85,000 in the Census and the 165,000 borne since, about 250,000 persons, was set up. Blood group records of 27,000 persons were linked to the database. This unique database was then used in various human genetics studies. This record linkage project aroused considerable international interest and was presented at meetings in the US and UK (Magnússon 1968). It was for this project that two magnetic disk drives were acquired, each removable disk holding 2 million digits.
An operational analysis of the herring industry was initiated in 1965. This involved collecting and analysing data on the location and catching of the herring, the landing ports, the landing facilities and capacity of the processing plants as well as distances from the fishing grounds. The idea was to optimise the whole system taking into account the overall economy and to be able to direct the fishing boats to the most advantageous landing ports from the point of view of distance to the ports, time for landing the catch and distance to the next fishing grounds. The owners of the fishing vessels and processing plants showed little interest, however, choosing to rely on the judgment of their skippers and, perhaps, guarding their interests as owners of the plants.
An operational analysis project to optimise the use of trawlers for supplying fish for freezing plants in North Iceland was undertaken in 1969 by the staff of the Computing Centre in cooperation with the Fisheries Association. This involved assembling and analysing large amounts of data on the Icelandic fishing fleet, the number and sizes of trawlers, their catch on each trawl, the fishing grounds and so on. Comparison was made between the profitability of trawlers of different sizes. The conclusion was that the most economical size of trawlers was 400 - 500 tonnes. This conclusion was, however, ignored by the authorities and the fishing industry and larger and less economical trawlers were built. (Benediktsson et al. 1969)
Due to the rapid changes in sea conditions around Iceland and the danger it entails for small fishing boats, an attempt was made to forecast the wave conditions around the coast of Iceland. This was actually prompted by the loss of a fishing boat due to unforeseen sea conditions. A program was developed for this purpose. The idea was to obtain data from the Meteorological Office in the afternoon, feed the data into the computer, run the program and broadcast the forecast at midnight. The program worked well in hindcasting but the problem was that it took the IBM 1620 24 hours to make a 3 hour forecast!
Already in 1965 plans were made to make a computer model of the Icelandic economy. In discussions with Professor Koopmanns, later a Nobel Prize winner in Economics, at a meeting in Iceland in 1965, it was pointed out that the Icelandic economy was dominated by the fishing industry and should therefore be relatively simple to model. Professor Koopmanns agreed and suggested that an attempt be made to make a model of the economic system. An economic modelling program package was obtained from the University of Vienna for this purpose. However, little progress was made in this area at the time, largely due to lack of interest by economists who were not ready to embark upon such an undertaking.
The computerisation of the real estate assessment process was initiated in 1966 and became a longterm project. (See the paper by Oddur Benediktsson et. al. at this conference).
In medicine there was interest in using the computer right from the beginning. The Cancer Register, i.e. a register of all known cancer cases in Iceland, was processed and analysed on the 1620. The Heart Association put all the records from the extensive epidemiological studies being undertaken by the Association on the computer for processing and statistical analysis.
Several institutes and companies started long-term projects in applied statistics using the computer. Of these the Agricultural Research Institute, the Marine Research Institute and the Meteorological Office were particularly active.
Actuaries from insurance companies used the computer to analyse the car insurance business among other things.
In the humanities a very interesting pioneering project was initiated on the IBM 1620.The project involved the frequency analysis of the words occurring in an Icelandic novel and led to a concordance project. Later similar analyses were made of several Icelandic sagas.
One of the earliest minicomputer based real time control applications in Iceland was installed in late 1974 in Landsvirkjun (The National Power Authority). The “Supervisory Control and Data Acquisition (SCADA) System” used was manufactured by Leeds & Northrup and was installed to control and monitor hydroelectric power stations and substations in the Landsvirkjun power system, which at the time covered the South-Western part of Iceland. The system, labelled the Conitel 2050, consisted of a computer based master station and several hardwired remote stations. The master station was based on a dual configuration of a Lockheed MAC-16 minicomputer. A similar system was installed at Reykjavik Municipal Electric Works. Both these systems were in operation well into the eighties.
Egill B. Hreinsson