Who Else Used LEOs?
One of the lesser known aspects of LEO's history is the extensive use of even the earliest LEOs by organisations other than J. Lyons & Co. The concept of the 'computer bureau' had arrived and Lyons were quick to capitalise on this interest as a source of revenue for the company, long before they started selling computers to others.
From the very first year in which LEO I started to become operational - 1951 - the computer was used by a wide range of organisations from government departments such as the Inland Revenue to the Met. Office to military and civil aviation companies.
Documents in the archive detail the following external LEO I users:
Ordnance Board (Ministry of Supply) - The first LEO computer was built by J. Lyons & Co. for themselves and ran its first full program successfully for Lyons on 30th November 1951, but even earlier that month Lyons were approached by the government's Ministry of Supply (the Ordnance Board) about using LEO for ballistic calculations. Lyons were awarded the contract for daily ballistic computations in January 1952 (source: the LEO Chronicle). Derek Hemy worked on programming for this job initially, although its content was kept secret for many years. It helped of course that the Ordnance Board had an office in Kensington, not far from Cadby Hall.
Sometime later it was revealed that LEO I was being used by the Ordnance Board to calculate the trajectories of missiles in connection with what became Black Knight. The Black Knight rocket was the United Kingdom’s first launch project and served as a test vehicle to study the effects resulting from re-entry to the atmosphere at high speed. The resulting data and lessons learned would be applied to the design of the Blue Streak missile.
Met. Office / Air Ministry - Mavis Hinds and Fred Bushby were using LEO I for meteorological calculations (weather forecasting) before the end of 1951 (source: Computer Story by Mavis Hinds). Using LEO I turned a 4-5 hour job when done manually into a 4-5 minute job on the computer. Leo fantl and Derek Hemy worked closely with Met. Office staff such as Fred Bushby and Mavis Hinds. The work with the Met. Office increased massively once card punches and readers were added to LEO I because the storage space was so restricted that data needed to be punched at every step within the enormous forecasting calculations required.
The LEO Chronicle also refers to working with the meteorological station at Dunstable on meteorological calculations (19th march 1952) and the Air Ministry (20th May 1952).
Inland Revenue - Lyons undertook a trial run and reliability test for tax tables in January 1952 (source: the LEO Chronicle). The test ran for 59 hours, 51 of which were useful, productive work, an amazing achievement less than two months after LEO I's first meaningful program run. The photograph of Ernest Lenaerts and Derek Hemy alongside a blackboard with reliability statistics and cricket scores on was taken during this test. The caption on this photo is 'Lenaerts and Hemy triumphant!".
Later, in April 1954, LEO was 'booked' to run the new tax tables for the governments' budget. LEO was ready to do so, but then the chancellor announced no tax changes at all in that budget... Ernest Lenaerts' notebooks for the period show how hard the LEO team worked to get things ready and then dedicates half a page to just the words 'No Tax Change!!". LEO's time to shine came in successive years though, as attested to by the various newspaper clippings in the archive (e.g. 64052 and 64165) and the LEO Chronicle which states "The P.A.Y.E. tables for 1955/6 are successfully produced on Budget night." for 19th April 1955.
Institute of Actuaries - in July 1952 Lyons agreed to to produce Annuity Tables using LEO I (source: the LEO Chronicle). LEO calculated "Joint Continuation and Last Survivor tables", though "due to a [human] misunderstanding" the first calculations were incorrect.
de Havilland Propellers - Lyons' quotation for using LEO I for guided missile calculations was accepted in September 1952 (source: the LEO Chronicle). Again, Derek Hemy was the programmer first involved and again he only later learned that he had been working on trajectories for the Blue Streak rocket. As Peter Bird records in his book, staff had to sign the Official Secrets Act and the computer room was closed off with red tape when sensitive jobs like these were run. As Bird also notes, Leo Fantl would have been the preferred person to work on the job but his Czechoslovakian heritage prevented security clearance.
De Havilland are still using LEO I in 1956 and the LEO Chronicle refers to the successful rocket trials at Woomera Test Range, Australia.
British Railways / British Transport Commission - LEO I is used for calculating the distances between every station and every other station, all 7,000 of them! (Source: LEO I and the BR job.) The intention was to identify the shortest route between each station, regardless of how impractical such a route might be. A 1957 edition of the Lyons Mail, Lyons' internal magazine, declared that this job involved 18 million calculations. The LEO Chronicle refers to the job as "the calculation and tabulation of the shortest distance between all pairs of stations in Britain".
Armstrong-Whitworth - LEO I worked on a "hyperbolic differential equation problem" and later a "linearised form" of the problem (1954) and "second order drag problem" and "second order lift problem".
Legal and General Assurance Society - Group Life and Pension Scheme job (1954), Annuity Tables for Australia (1955) and Group Pension quotations (1956).
British Road Services (the nationalised road haulage industry) - LEO I worked on tables of tariffs and rates.
Ever-Ready - LEO I worked on van sales stock control in 1955, then salesmen's stock control and sales statistics in 1957.
Royal Bank of Scotland - although there are scant records of LEO I actually doing the job, the LEO team were actively involved - and regularly quoting for - how customer accounts could be controlled using a computer. This was a very early proposal in using computing in banking. Later, LEO would be involved in BARIC.
Ford - the motor company's payroll ran on LEO I from 14th November 1955. The first run involved 700 employees, by 1956 LEO I is running the job for 6,000 employees and by May 1957 for 10,300 employees.
Smiths Aircraft - LEO I ran its first job for them in January 1956.
Alliance Assurance Co. - LEO I worked on Group Pension Scheme quotations (1956).
Attwoods Statistics -
C.A.V. - Injection Systems job (1956)
Office Management Association - Clerical Salaries Analysis (1956)
Imperial Tobacco - pension fund job (1856)
Dairy Farmers - Dairy Exhibition job (1856)
Stewarts & Lloyds (steel) - payroll pilot run begins May 1957 and LEO I works on their "pipework stress and anchor problem" in July of that year.
North Thames Gas Board - central stores job (1957).
The second LEO built was also developed by J. Lyons & Co. for themselves. Lyons always intended to build two computers for reliability reasons but by the time the first LEO was declared 'finished' in 1954, they were able to make considerable changes to the computer's design and so the second generation of LEOs was begun. By this time, LEO I had many external users so Lyons decided to build LEOs for others too. LEO II/1 became operational in 1957.
Records are not currently clear on when individual customers' work was switched to LEO II/1 from LEO I, but by 1958 many organisations external to Lyons and LEO Computers were regularly using the computer bureau function. These included Kodak (payroll), Tate & Lyle (payroll), Stewarts & Lloyds (ore digging), Dunlop (sales accounting) any many others. Many of these organisations subsequently place orders for LEO computers of their own.
This exhibit has a reference ID of CH70221. Please quote this reference ID in any communication with the Centre for Computing History.