Westwood Works 1903-2003
Probably the first mechanical calculating machine was the abacus, also called a counting frame, a calculating tool for performing arithmetical processes, often constructed as a wooden frame with beads sliding on wires. It was in use centuries before the adoption of the written Hindu-Arabic numeral system and is still widely used by merchants and clerks in China, Japan, Africa and elsewhere.
Another precursor of the computer was the Comptometer. Patented in 1887 by Dorr Felt, a US citizen, the Comptometer was a type of mechanical (or electro-mechanical) adding machine designed primarily for adding but later capable of carrying out division, multiplication and subtraction. It was the first adding machine to be driven solely by pressing keys.
Comptometers were developed for many specialised purposes with from 30 to over 100 keys arranged in horizontal and vertical rows.A skilled operator could process data much faster than with an electronic calculator as, using as many fingers as required, all the digits of a number could be entered simultaneously. Comptometers were used well into the computer age, until the 1990s for some specialist applications.
Comptometers were used at Baker Perkins during WW2 - see here.
The quest to produce an efficient calculating machine goes back at least to the beginning of the 17th century and the work of Blaise Pascal. A child prodigy, born in 1623, Pascal was a brilliant mathematician who, In 1642, and not yet nineteen, in an effort to ease his father's endless, exhausting calculations, and recalculations, of taxes owed and paid, constructed a mechanical calculator capable of addition and subtraction to help his father with his work.
Jacques Vaucanson was born in Grenoble, France in 1709 and first made his name with the construction of intricate automatons. Appointed by Cardinal Fleury, chief minister of Louis XV, as inspector of the manufacture of silk in France, he was charged with undertaking reforms of the silk manufacturing process. At a time when the French weaving industry had fallen behind that of England and Scotland, Vaucanson developed automation of the weaving process. Creating, in 1745, the world's first completely automated loom using punched-cards - a technology that, as refined by Joseph-Marie Jacquard (1752 – 1834) more than a half century later, would revolutionize weaving and, in the twentieth century, would be used to input data into computers and store information in binary form.
Jacquard was a weaver, and he developed a loom that worked different coloured threads into his fabric according to patterns that he punched into cardboard strips. This was the forerunner of the punched card. By passing compressed air through the holes, the loom would feed certain threads into the warp and weft of the fabric and patterns would result. The Jacquard loom is still in use to this day.
Charles Babbage (1791 – 1871) was an English mathematician, philosopher, and mechanical engineer who originated the idea of a programmable computer – his famous “difference engine”. Parts of his uncompleted mechanisms are on display in the London Science Museum. In 1991, a perfectly functioning machine was constructed from Babbage's original plans. Built to tolerances achievable in the 19th century, the success of the finished engine indicated that Babbage's machine would have worked.
The ideas that Babbage left behind lay fallow until 1889, when Herman Hollerith had patented his ideas for a punched-card – based system and had taken a copy of his prototype to the Universal Exhibition at the Eiffel Tower, Paris, where he won a Gold Medal for his Electrical Tabulating System.
The United States Census bureau was taken every decade. The processing of the information then took a further seven years before results could be published. In 1898, the United States Census Bureau offered a prize to anyone who could devise a system by which the results of the ten-year census could be speeded up. Doctor Herman Hollerith who happened to be an employee of the US Census authorities took up the challenge.
Building on the ideas of those who had gone before, using a punched card and an electrical sensing mechanism which could count, Hollerith devised a system to transfer all the census details onto cards, using his sensing mechanism (called a pantograph) to read them and his counters to record results.
The pantograph was a gadget which had a whole lot of spring loaded feelers which, when operated, would align themselves with an equal number of little dishes filled with mercury. If the card being sensed had a hole, the feeler would make contact with the mercury and complete an electrical circuit. If there were no hole, of course, there would be no contact. If a circuit were completed, some sort of arithmetical calculation would result. The census of 1890 was completed in two years, and Hollerith won the prize. In later years, the original machines were improved in various ways, and though the punched card itself remained, and still is the same size as the original, its ability to record data increased from about 24 to 80 items of information.
In the formative years, a gentleman named James Powers - a Russian-born mechanical and statistical engineer who came to the United States in 1889 - assisted Hollerith. The two disagreed on the preferable sensing methods, with Hollerith favouring electrical sensing while Powers favoured mechanical sensing. This disagreement (which was quite amicable in its character), ultimately led to the formation of two of the giants of the computer industry (See The birth of the Computer Business below).
Hollerith used an elongated rectangular hole in a card that could carry 80
columns. As a result, Hollerith machines needed to use a system which was much
more precise than that used by Powers. The Hollerith hole-making device called
a "die and stripper" was therefore a high quality piece of precision
IBM and BTM worked with one another for many years, with IBM serving the Americas and BTM the rest of the world. IBM concentrated on the development of tabulating and punching equipment, while BTM developed sorters, collators, gang punches, reproducers and interpreters. The two companies shared their developments until the 1940s when American anti-trust laws forced them to compete with each other and a bitter rivalry ensued.
Following business disagreements with Herman Hollerith, the US Census Bureau hired James Power to build punch card equipment for the 1910 census. Powers’ punch card featured 240 keys. Each key corresponded to an item on the census questionnaire. Pins were set as each key was pressed. All columns were punched simultaneously when the card was finished. The particular advance was that data could be entered for the entire card before the 90-column card was physically punched with round holes, which let census workers correct incorrect data before punches were made. Once the data was complete and correct, the simultaneous punching occurred.
Powers left the Census Bureau in 1911 after the government gave him the right to patent his machines. Powers then started Powers Accounting Machine Company (also called Powers Tabulating Company) in Newark, N.J. He moved the company to Brooklyn in 1914. There he sold his punch, sorter, and a printing tabulator. Patent and licensing issues handicapped the company, and it nearly closed during the 1921 recession.
Powers’ Accounting Machine Company remained the principle competitor to Hollerith’s Tabulating Machine Company. Powers used a 21, 45 and a 64-column card. The holes were round, and thus they were not able to fit 80 columns into their large card that was exactly the same size as Hollerith’s.
As mentioned above, the disagreement between Herman Hollerith and James Powers led ultimately to the creation of IBM (International Business Machines) (in America) and BTM (British Tabulating Company Ltd) (in Europe) both of which developed along the electrical sensing route. Later, Hollerith had merged with two companies to become CTR (Computing-Tabulating-Recording) and, under the leadership of Thomas J. Watson, CTR had become IBM (International Business Machines).
Remington Rand acquired Powers’ company in 1924, Powers’ company becoming the tabulating machine division of Remington-Rand and competing directly with IBM, Hollerith’s successor company.
Powers’ development of a mechanical sensing system led, in 1932, to the formation of a company named Power Samas. (The establishment in Paris, in 1914, of the International Time Recording Co. had marked the start of the information processing industry in France. Eight years later, (Societe Anonyme de Machines a Statistiques), was established in Paris to market Powers’ machines). The French company was so successful that its name was incorporated into the European company, Power Samas). Power Samas developed along similar lines as the other companies, but was never compatible with them.
Power Samas equipment was known commonly by the nickname “Acc and Tab,” an abbreviation for Accounting and Tabulating. It captured about fifty percent of the British market against their primary rival, British Tabulating Machine Company (BTM), which was licensed to provide IBM products.
In the late 1950s, BTM merged with Power Samas to become International Computers and Tabulators Ltd (or I.C.T.) with BTM holding 62% of the equity and Powers 38% in the new company. For a while, after the Power Samas merger, the mechanical sensing system was continued by the new company, but was phased out in favour of the Hollerith type system. In the late 1960s, I.C.T. changed its name to International Computers Ltd (ICL) when it merged with English Electric and Plessey.
It is not clear when Baker Perkins first installed a punched-card system but it is likely to have been during the late 1930s. It is known that Power Samas supplied the equipment. By January 1947, this equipment was nearing the end of its useful life and was being replaced by Hollerith equipment.
In July 2005, Hugh Brenton wrote a chapter entitled “The Computerisation of Requisitioning at Baker Perkins Ltd” for Jim Farrow’s book “The History of the CPO”. (Copies available through this website). In this he referred to the “Power Samas ” and “Hollerith” punched-card systems that had been used in the Accounts and Costing Departments prior to the introduction of computers.
“My cousin Paddy Adams (nee Pearce) started work in the Punch Room in January 1947. At that time the Power Samas machines were still in use but were nearing the end of their working life. Paddy was therefore only involved with the Hollerith equipment which was being installed at about the same time as she started work”.
Hugh, who joined Baker Perkins as a Commercial Trainee in 1961, (see Trainees in Action), spent part of his training in the Cost Office and the notes that he took at the time have been used to explain some of the uses of the Hollerith punched-card machines at Baker Perkins during the early 1960’s.
The Hollerith cards used were 80 columns wide and each card was punched by a first operator and then verified by a second operator (to eliminate any punching errors). The cards could then be sorted or tabulated by using the appropriate Hollerith machinery.
(See also The CPO).
Jim Farrow’s book (referred to previously) gives full details of the requisitioning process and explains how a set of works paperwork was printed for each component or assembly to be manufactured. Each set of paperwork carried a unique identifier known as the “piecework contract number” and the documents printed included three for use by the Cost Office – these were a Cost Requisition Card, a Material Issue Card and a Work-in-progress Card. Two of these (the Cost Requisition and the Material Issue) were 80 column Hollerith cards ready to be punched.
On initial sorting of the works paperwork in the CPO the Cost Requisition Cards were separated and sent to the Cost Office. Their receipt was recorded and they were passed to the Hollerith Section to be used as a “Filing Master”.
On the issue of the first item of material for any piecework contract (pwc), the Material Issue Card and the Work-in-progress Card were sent to the Cost Office. The Work-in-Progress Card was filed and the Material Issue Card was married-up with the Cost Master Card for the same drawing number.
The Cost Master Card had been created previously using information from the Job Specification to give the standard material, labour and overhead costs for “one-off” the component or assembly concerned.
The Material Issue Card and its Cost Master Card were then sent to the Hollerith Section where the Material Issue Card was punched with the batch quantity and the pwc number. If the batch quantity was greater than one, additional Material Issue Cards would be copied until the number of cards equalled the batch quantity e.g. a 6 off batch would generate 6 material issue cards.
The resultant material issue cards were then punched with the material, labour and overhead costs. These costs were copied from the Cost Master Card but with the setting time costs reduced by dividing them by the batch quantity e.g. for a 6 off batch each material issue card would carry 1/6th of the set time cost.
In addition to punching the Material Issue Cards, the Hollerith Section also used them to tabulate a material issues valuation which would be used at regular intervals for accounting purposes. After this, the punched Material Issue Cards were returned to the Cost Office to be held in storage trays known as the Stock File.
As each operation of a pwc was completed the operation card (time ticket) was sent to the Cost Office from the CPO. The time ticket was firstly used for bonus calculation and then forwarded to the Hollerith Section to be recorded as “complete” in the work-in-progress file held with the Cost Requisition.
The final completion of a pwc was notified to the Cost Office by receipt of
the Progress Record Card from the CPO (plus the Stores Sheet for assemblies).
This caused the Cost Requisition, and all the other cards filed with it, to be removed from the Hollerith work-in-progress file and forwarded to the Strong Room for archive storage.
The cost of each assembly was “built up” by the Cost office on receipt of the completed Stores Sheet. The Material Issue Card for the assembly was withdrawn from the Stock File and then the Material Issue Cards for each item of “piecework stock” listed on the Stores Sheet were also withdrawn. The assembly card would give the material and labour costs for the assembly itself and the “pick-up” cards would give the material and labour costs for the piece-parts required. If the assembly picked up more than one of a piece-part the number of cards withdrawn would be equal to the quantity required.
All the cards withdrawn were then sent to the Hollerith Section who used them to compile a tabulation showing the material, labour and overhead values for that assembly. This tabulation was used to create an Assembly Cost Card which was filed in the Stock File until it was “picked up” itself on a higher level assembly.
The material issue cards that had been used to produce the assembly cost tabulation were then destroyed.
Thus the various assemblies were cost-built until they were in turn picked-up on the key arrangement to give total material, labour and overhead costs. The stock account 13 electrics (listed on the Electrical Data Sheet) plus a percentage allowance for stock account 16 hardware (nuts and bolts etc) were then added manually to give a total ex-works product cost.
The costing process was therefore kept in step with actual events in the factory by use of the Hollerith punched-card process - this was a very sophisticated system for its time, preparing the way for the electronic computer systems that were to come later.
Hollerith punched-card systems were used for other applications within the company such as wages and bonus calculation. No detailed information about these areas is currently available but it is hoped that another ex-employee will be willing to help. In pre-computer days, "Hollerith" cards were punched with information about an employee's clocking-on/clocking-off times and these cards were passed through a form of calculating machine - pre-loaded with each person's pay rate, etc - which worked out their weekly or monthly pay. As this was in the days before direct payment of wages/salaries into bank accounts, this information was then passed to the Salaries and Wages Department where the appropriate sums were put, in cash, into small brown envelopes to be handed out on "Pay Day". (See Departments at Westwood - "Salaries and Wages Department" and Clocking-On).
Baker Perkins was an early user of computer-based business systems with the installation of an English Electric Leo KDF 6 Data Processing system in 1964. Used at first for payroll work, producing wage and bonus slips, its operation was extended to include factory production control, costing and general accounting procedures. A second machine was installed at the end of 1969. The space taken up by these computers – nearly two floors of the 1933 multi-storey block - would, today, house many hundred PCs and the computing power of each was perhaps not even equal to one of today's desk-top machines. This is covered in The CPO – The 1960s – Computerisation. Photographs of the first and second English Electric Leo computers are to be found in – Life at Westwood Works – In the Offices – 1900-1969.
Inevitably, this spelled the end of the use of punched-cards. The Hollerith Department finally closed in 1964.
Christine (nee Spall) Aitken joined Baker Perkins in 1958, starting in the Commercial Archives before becoming a typist in the Planning and Ratefixing Department. Christine recalls - "When Baker Perkins first installed a computer, it was used to process the Staff payroll and job specifications for the Planning and Ratefixing Department. Originally, the payroll was processed in the Accounts Department by girls inputting the information onto Hollerith machines. Typists in the Planning and Ratefixing Department produced job specifications for the shop floor which were typed onto paper with a carbon backing. These were then put to the Print Room for copies to be run off.
In 1964 , the girls from Hollerith and the typists from Planning and Ratefixing were transferred into the Computer Department where they each had a machine that punched out holes in a tape. This tape was then fed into the enormous computer where all the data was processed and printed copies produced for payroll and planning and ratefixing. Joan Baker was our supervisor.
I worked there until 1968 when I married and moved away from Peterborough. Maurice Seago was my boss whilst working in Planning and Ratefixing and Mr Ramsden was head of the Computer Department".
The Computer Punch Room girls circa 1964. Christine is in the centre of the front row.
Mary Baxter worked at Baker Perkins from August 1965 to June 1993. Mary recalls -
"For the first year of my employment there I was in Final Accounts and Invoices Inwards in the Accounts Department. I then moved into the Cashiers Department until October 1973 when I became a Trainee Programmer in the Computer Department ending up as a Senior Programmer."
Mary remembers some of the "behind the scenes" activities:
"The computer runs 24 hours a day, therefore the operators worked a shift system. I don't recall whether the night shifts always operated with two people on duty or whether that was put in place after this incident but there was one night at least when there was only one operator on duty. As you can imagine, there was no-one else in the whole of the offices all night. On this night the operator collapsed and became unconscious but it was very fortunate that not too long after this happened one of the security men on duty made his rounds, which included calling in on the computer department, and found him. All was well and he made a full recovery as far as I recall. After that there were always two operators on the night shift. This was I think in the late 1970's or maybe the early 1980's."
"Because computers and their associated equipment - the large stuff that is - generate a lot of heat they need air-conditioned rooms to operate in. This meant that the computer department as a whole was quite often the most up-to-date office in the building with regard to mod cons and decoration and furnishings. On the rare occasions that the department was moved it was all done again in the new location. I think it possible that some people, or some departments, resented this a bit when they felt they were left working in less than ideal conditions. This may have been more so in the earlier days, 1970's and early 1980's. Perhaps you could ask around to see if this was the case."
"In the 1970's for a period there was a fairly regular Bridge session at lunchtimes, in the days when we still got an hour for lunch. Some people took part and others stood around and watched the play. Some of the players were very good and I was told that Ollie Gilbert was one to watch because he had the ability to remember all the cards that had been played and therefore knew what was left! I had a go myself once or twice but never quite mastered it. Perhaps I just needed more practice!"
All content © the Website Authors unless stated otherwise.