memory cards for the very first AXE exchange, in Sodertalje, tested in 1976. Photo: Ericsson’s archives at the Centre for Business History
telephone station system turned Ericsson from a third-tier telecommunications companies to global leadership and has been sold for hundreds of billions.
Spring 1972 is LM Ericsson’s management face a difficult choice. Which system for local telephone exchange you should invest in? It largely already fully developed AKE, or the new AXE, which is only available in the form of reports and drawings from the development company Ellemtel
Choosing AXE will mean higher development costs and the time it takes before a complete system are clear. While it promises significant benefits in the long term.
It is a difficult choice for a company whose main product is still relay-based switches. One industry where one is accustomed to products manufactured from sheet metal and copper wire, not the processor chip and program code. At last fall anyway elections on AXE.
It was not many years before it became clear that Ericsson chose right. AXE was completed in time to be involved in the first major tender competitions and did not lose a single one of these. The system bar Ericsson from electro-mechanics into the fully electronic and digital technology. I? Today, four decades later living on in Ericsson AXE products for both fixed line and mobile.
story began in the fifties USA at Bell Laboratories. Computers, or mathematics machines, they still called, was in vogue and would be used for everything. At Bell realized that they too could control switchboards. They took out a patent on the idea and developed a prototype, which became operational in 1958.
In Sweden the Swedish Telecom early inside the similar thoughts. 1963 the plant its first prototype of a computer-controlled switch, Test-1, finished. Meanwhile, plans began to seriously use the new technology to take firmer shape. At this time manufactured Telecommunications Authority much of its equipment in-house, while LM Ericsson mainly focused on the export market. A large part of Ericsson’s customers in Latin America and Asia, but has now more difficult to compete in the Western European and North American markets.
In the 1960s, built the LM Ericsson and Swedish Telecom was his computer gear. Both, Ericsson AKE-12 in Tumba and Televerkets A 210 in Storängen, were completed only a year later than planned.
One important lesson was that the programming of the turnouts computer system was much more difficult than anticipated. The programs became opaque, corrections in a program section created unforeseen consequences for other parts of the system. Application volumes were larger than expected, so much larger that the stations’ capacity decreased, because the programs seized more memory than expected.
Telecom and LM Ericsson was therefore natural to pool their resources. In 1970 they formed the joint development company Ellemtel.
During the first two years makes Ellemtel different systems of investigations and try many different suggestions.
? – At the bottom was the experience of the parents’ past activities, told one of AXE’s main architects, Bengt-Gunnar Magnusson at Ellemtel, in an interview with New Technology 1978.
-? Televerkets A 210 for example, had divided the hardware of the control system on a central processor and regional operating units. The same principles will return in AXE’s hardware with central and regional processors, although the latter are intelligent and control units were stone stupid and did not pass any signal processing.
From Ericsson AKE project came the solution to dealing with the “big program volumes.” Meanwhile, with the final phase of work on Tumba station had been the development of AKE-13, a major interchange station.
Software structure is threatened to be even more confusing than Tumba station. One of the employees was Ivar Jacobson. He proposed to define the structure of the system is needed, which is a block diagram with defined interfaces between blocks, where each block would represent a telephony function.
Ivar Jacobson was the first to develop the concept of function blocks became his colleague Göran Hemdal who picked up all the opportunities it afforded. He attended AKE 13-work and was only a stubborn opponent of the function blocks. Hemdal put a lot of work to prove that the idea was flawed, partly because it created large volumes of applications, but instead came to the conclusion that it was a substantial advantage. He is probably the single person who meant the most to the development of AXE, although he did Ellemtel back in 1978 and went to the worst competitor ITT. The basic ideas he drove forward in the AXE project was split into function blocks would be combined with a specific programming language and also affect your computer’s internal structures.
Goran Hemdal answered himself for the design of programming language Plex and much of the practical programming work.
After two years of work, and many rejected the idea, put Ellemtel the spring of 1972, the proposal for the AXE. The system was based on a total modularization that went through even in the mechanical construction method.
Then there arose a doubt in LM – was this the right way? To further develop the AXE to the finished product would certainly give technical advantages, but required including a brand new computer. Retaining the old AKE computer would require less development and more quickly provide a complete system. There was a risk that LM would come out too late on the market if they scored at AXE road.
arguments were heavy on both sides, and Björn Lundvall, LM’s then CEO, is quoted as saying that “at last we shall have to toss heads or tails of it.” The Telecommunications Authority was LM’s certainly trying. Telecommunications Authority had largely moved across its expert staff to Ellemtel, and was set on a new system.
In May 1972 came the green light. AXE project began in earnest with a very tight timeframe. During the main part of the development work took 150 to 200 of Ellemtels employed. Added a further borrowed techniques from the parent companies.
1973, the project was changed conditions. To AXE could hang with must be able to offer a digital electronic “group of voters” instead of the analog relay selector that was included from the beginning. Ericsson subsidiary in Australia had made fundamental studies of electronic voters and had developed a prototype.
introduction of the electronic group selector was the first real test of funktionsmodulariteten. The amendment was carried out without the rest of the system needs to be corrected.
timetable of 1972 aiming for a complete system in operation by the spring of 1976, just in time for LM Ericsson’s centenary. It was optimistic on the high side. True, the company’s new CEO Björn Svedberg showcase photos from the first Swedish AXE station in Södertälje during the birthday celebration, but the switch was by then far from finished.
Meanwhile, a growing number of competitors announced their own computer-controlled local exchange systems, ITT with its Metaconta. The rumor went that LM Ericsson was disastrous for and not managed the transition from electromechanical to electronics.
February 1974 asked Dagens Nyheter exchange page uneasily “Will LM Ericsson 1980s Facit?” The same concerns echoed in an article from October 1975.
“Some of the competitors and Telecommunications journalists who spoke off the record in Geneva said that LM’s AXE system is a few years after its competitors (eg Siemens, ITT, GTE, and Japanese).”
oil crisis in the mid-1970s and the subsequent recession played LM Ericsson in your hands. Many governments had delayed their orders for new telephone system. These included Australia, who planned a major expansion and modernization of the telephone network. The land had previously been a faithful Ericsson Client, but in a previous contract had chosen another supplier.
demand started to come back, just in time Ellemtels technicians successfully started AXE station in Södertälje, so it was inaugurated in March 1977. It was time for the Australian telecommunications management would get to see it in operation before where did their system choices.
In September 1977, news came: Australia wanted AXE. It became the first in a series of large export orders. The most spectacular was featured on Lucia Day that year. A consortium of Bell Canada, Philips and Ericsson had received orders for a total modernization of Saudi Arabia’s telecommunications networks. Only Ericsson’s share was worth 4.5 billion, larger than any previous Swedish export orders.
success was a fact.
/templates/v3/img/icons/Mail-envelope.png){kind=icon}
