Ferranti-Packard

Ferranti-Packard was the Canadian division of Ferranti's global manufacturing empire, formed by the 1958 merger of Ferranti Electric and Packard Electric. For several years in the post-war era the company underwent dramatic expansion and repeatedly "almost" became a major computer supplier, but eventually shed various divisions and returned to becoming an electrical grid supplier once again. The company was purchased in 1998 by the Austrian company, VA TECH.

Packard Electric had first been set up in 1894 in order to supply transformers during the Niagara Falls hydroelectric developments. Ferranti Canada had first been set up in 1912, acting primarily as a sales and distribution arm for their British designed electrical products.

Prior to World War II, Canada's economy was primarily agricultural. While this allowed it to raise a fairly large army with relative ease, it also meant that it was unable to arm itself. C.D. Howe, Minister of Armaments, started an ambitious plan to heavily industrialize the country, turning it almost overnight into what is today a G8 nation.

Canada had entered World War II completely unprepared, and in the post-war era decided they would not allow this to happen again. However, as the art of war turned increasingly technical, it was clear that Canada did not have the wherewithal to support a full program of research on its own. In 1947 the Defense Research Board (DRB) was formed, and sent out a letter outlining their ideas for sharing research between the armed forces, industry and academia.

The letter made its way all the way to the desk of Vincent Ziani de Ferranti, the then-current CEO of the family-held British company. At the time Ferranti in the UK was involved in a similar commercial/academic development project with Manchester University to build the Manchester Mark I computer, so it seemed that their Canadian division would naturally be able to do the same. In October 1948 he flew to Canada to meet with the DRB. He was disappointed to learn that the DRB did not have the financial resources to fund any sort of program, but Ferranti remained interested even though it appeared the only way to work with the DRB would be for free.

Just such a project started soon after, when word of the Ferranti meeting reached Jim Belyea, a researcher in the Canadian Navy's electrical labs. He had been proposing a completely automated system for ships to pass around tactical data from radar and sonar, to help organize the defence of a convoy under attack by submarines. Belyea presented his ideas to Ferranti, who agreed to start development of the technologies needed. By 1950 they had successfully developed a PCM-based radio system for passing digital data between ships, and the DRB started to become very interested. Full-scale development of the system, known as DATAR, started in February 1951 and underwent trials in late 1953. However the cost of developing a production version was well beyond what the Canadian Navy was able to afford. They attempted to sell the system to the United States Navy, but they were uninterested and the Canadians were forced to end the project. Somewhat ironically the US Navy would later decide they needed just such a system, having a rather unhappy experience with their Naval Tactical Data System in 1958.

During this period, some time in 1951, Ferranti Canada also considered commercializing the University of Toronto's experimental UTEC computer, which seemed considerably less complex than the Mark I being developed in England. This effort also ended ironically, when in 1952 the University purchased a surplus Mark I originally intended for the UK's nuclear weapons program which had suffered massive budget cuts with a change of government.

In 1956 the company received a contract from the Canadian Post Office to develop an electronic mail sorting system, which they delivered later that year. The system used a hard-wired transistorized computer that stored a table of postal codes on a magnetic drum. Operators were presented with envelopes and typed in the postal code, which their typewriter printed onto the envelope as a bar code in fluorescent ink. The sorting system would then read the bar code and automatically route it sort it. The system was a complete success.

In fact it so impressed visitors from the US Post Office that they decided they needed one of their own. They also decided to develop their own system instead of simply buying the Ferranti one, delaying their entry into automated sorting until 1960. A Canada-wide expansion using Ferranti's existing system soon ended in the 1957 election, whose main issue was rampant Liberal spending, including Ferranti's "million dollar monster".

Oddly the system was later adapted for cheque sorting by the Federal Reserve Bank in New York, who took delivery of an almost identical machine in 1958, based on reading MICR digits instead of bar codes. This MICR-based concept was originally developed by SRI in 1952 for their similar ERMA project, but they did not manage to actually deliver their machine until 1959. There was some talk of developing the Ferranti system into a commercial line, but it became clear that as general purpose computers fell in price, a single mass-produced model would soon be able to outperform a custom built design, even on cost. Honeywell started shopping around just such a system at about this time.

Another brush with success came in 1959 with the ReserVec on-line reservation system developed for Trans-Canada Air Lines. This product suffered from stiff competition from IBM's SABRE system in the US, but there was nothing similar in the UK, whose own airlines were in the market for such a system. Apparently due largely to not invented here problems, Ferranti in the UK decided to develop an entirely new system to fill this need instead of using the Canadian version.

But by this time Ferranti's UK computer divisions were themselves in turmoil. Their attempt to commercialize the Atlas design was dragging on, and meanwhile sales of their older Mercury were drying up. In order to address this, as well as move into a new market segment, they decided to launch a newer system aimed at the low-end of the market. The result was the Ferranti Orion, which used an entirely new circuity system known as "Neuron". This proved to be somewhat of a disaster, and only 12 Orion I machines ever shipped (the first went to AB Turitz and Co., of Gothenburg, Sweden, in March 1963).

Meanwhile Ferranti-Packard decided they should set up production for the Atlas machine as well, but after successfully securing loans from the government they were astounded to learn that the UK division refused to allow them access to the design. Many of the company's engineers resigned in disgust, although some were later convinced to stay on. The UK division then asked several Canadian engineers to move to England in an attempt to re-engineer the Orion based on ReserVec's transistorized circuits. Known as Orion II, the project ran in parallel to the original Orion for some time. (Orion II was much more successful, both technically and commercially, than its predecessor. Nearly 40 machines were delivered by the end of 1964.)

With the experience gained during Orion II the engineers returned to Toronto convinced that ReserVec's design sold into the Orion's marketplace would be a commercial success. Once again approaching the Federal Reserve Bank, they proposed to build a new machine to replace the earlier post office-derived system. The Bank accepted the proposal, and work on what would become the Ferranti-Packard 6000 started in late 1961.

In order to differentiate themselves from the numerous models in this performance range, the FP-6000 would directly support multitasking, then known as "multiprogramming", as well as be highly modular. The prototype machine was completed in 1962, and was delivered to the FRB took place in early 1963. Further sales proved difficult however. One was purchased by the DRB's station in Dartmouth, Nova Scotia, another by the Toronto Stock Exchange, one by a Scottish research facility, and a final machine by Saskatchewan Power, bringing the total to five machines.

Meanwhile, unknown to Ferranti-Packard, Ferranti in the UK had decided to cut their losses and exit the commercial computer business. In early 1963 they approached International Computers and Tabulators with the proposal to sell off their commercial division. Perhaps unsurprisingly, ICT found the proposal unattractive. When they learned of the FP-6000 their attitude changed, and they eventually agreed to the purchase under the stipulation that rights to the FP-6000 would be transferred from Canada. The FP-6000 then became the basis for ICT 1900-series of machines, which eventually sold into the thousands.

After the ICT takeover, Ferranti-Packard proposed that they manufacture several models of the 1900-series, as well as serve as a gateway into the North American market. ICT, however, was interested in Europe only. Most of the FP-6000 engineers soon left the company, forming ESE Limited. The programmers founded I. P. Sharp Associates in December 1964.

One of the Canadian company's last developments was the flip-disc display, which creates a large display out of a grid of small disks, painted black on one side and some bright color (typically yellow) on the other. Small magnets on the back of the disks flip them over if the wires running behind them are powered. The UK headquarters gave the invention little note, but it became somewhat successful in spite of this, and HQ eventually used it as a way to try to sell off the Canadian division in the 1970s. Today these displays can commonly be found in outdoor use, notably on highway signage and in the automotive application of destination signs for Public Service Vehicles.

With the electronics division empty, Ferranti-Packard was once again a major electrical vendor only. Over the years many other specialty divisions were sold off or closed, and eventually all that was left was the original Packard transformer division. It was this that VA TECH purchased.

References

* [http://ieee.ca/millennium/fp6000/fp6000_datar.html From DATAR to the FP-6000]
* [http://rogerdmoore.ca/JOUR/ TIME-SHARING ON THE FERRANTI-PACKARD FP6000 COMPUTER SYSTEM]
* [http://ferranti-orion.co.uk/ The Ferranti Orion Computer System]