For something that started life as a humble switchboard, the PABX or Private Automatic Branch Exchange has come a long way. It may not be the most glamorous market, and unlike the computer market is far from volatile, but it is worth some $2,900m a year in Western Europe, according to Logica Plc, and is set to grow to $3,370m by 1992. But in view of the approaching spectre of Integrated Services Digital Network, ISDN, and the introduction of Centrex, is life in the PABX market likely to change – will the players themselves be different, and how will the kinds of systems being developed be influenced by these newer technologies? Saturated According to industry watchers, those adjectives that have applied to the sector in the past will continue to best describe it – healthy and steady, with a volume growth of 24% expected between 1987 and 1992, the equivalent of a value increase of 16% leading to that $3,370m figure. Although the market is saturated, sales will not be unduly affected, with replacement systems dominating shipments over the next few years. In 1987, for instance, replacements amounted to a whopping 69% of total shipments, a figure expected to rise to 78% by 1992. One of the main factors affecting the purchase of replacement systems is ISDN, which has encouraged many users to swap their ageing analogue PABXs for digital systems. It has also resulted in a demand for products that can demonstrate speech and data integration. Many customers seem happy to buy new systems, despite the fact that many of the advanced features cannot yet be used, and the fact that many of them will require substantial upgrades to be compatible with the CCITT standards. Actual integration of speech and data has, reckons Logica, been slower than expected, because of administrative-type problems like ascribing responsibilities within the user organisation as much as the fact that, so far, prices have been high. Despite this, the number of digital extensions is expected to have doubled to around 30% by the end of the decade. What the Logica study found was that users are introducing a two-tier data communications strategy, so that if the data connection requirement is low, and the provision of separate cabling is difficult to justify, data will be integrated with voice and switched through the PABX, while high volume requirements will be met with local area network connections. This is a conclusion that Stephen Timms, principle consultant on an Ovum Ltd report, ISDN: Customer Premises Equipment, also comes to: While PABXs certainly have a long-term future ahead of them, it will not be as the main switches for data. Large sites will continue to have separate data switches, while the smaller outfits will really make use of voice-data integration. There have certainly been changes for the manufacturers – for the costs of developing systems which can meet the transmission demands of high-volume local area network-type data communications are very high. Logica reckons that development costs for a new range can often be as high as $100m, and that to offset such costs manufacturers are looking at possibilities like entering new country markets, or embarking on much more merger and joint venture activity. Just a glance at Dataquest’s league table for Western Europe shows how the mergers of the last two years have shuffled the positions in the league table: Alcatel NV, for instance, has booted Siemens AG into second place, boasting a total of 1.2m lines, while Telenorma AG – comprising Robert Bosch GmbH and Jeumont Schneider SA – steps into the number three position. GEC Plessey Telecommunications Ltd comes in fourth, followed by L M Ericsson Telefon AB, Northern Telecom Ltd and Mitel Corp, with a mere 300,000 lines each. Interestingly, AT&T Co and IBM Corp are absent from the league as both are currently seeking approvals for their own digital products. Where there has not been a merger, there has been some kind of distribution or marketing deal, such as the one between Italtel SpA and Plessey Co. So while there will be a few Europe wide su
ppliers promoting a range of products over all sizes and classes of PABXs, there is also expected to be a few small specialist firms going after specific areas. However the manufacturers have been pairing themselves off, ISDN has certainly affected PABX design – particularly of larger systems with over 100 extensions, and has led to the imaginatively named ISDN PABX – basically a PABX with digital switching and a CCITT or national standard interface to the public network. According to research done by Ovum Ltd, there are currently systems on offer from 13 suppliers, with a number of other products known to the researchers but not commercially available. It is estimated that the cost of ISDN functionality will add around 20% to the price of a large PABX in 1989-90, but that the premium will disappear within five years. One of the major influences of ISDN on PABX design has been on a country-by country basis, depending on the access type preferred by that particular area. For instance, in the UK and the US, the emphasis is on corporate networking, where primary access is the first requirement, while in France and West Germany a very different picture emerges. In the US, AT&T and Northern Telecom are both carrying out primary access trials with System 85 and Meridian SL-1 exchanges respectively, both the Deutsche Bundespost and France Telecom have launched initial ISDN services offering basic access, with suppliers falling into line. Pay for a full 30 A further implication of the introduction of primary rate access was highlighted by Terry Wright, an analyst with Dataquest: Basic access defines the user access to the network, while primary access defines the connectivity between PABXs, a PABX and a computer and the PABX and the central exchange, and while Basic Rate Access is just one channel, Pirmary Rate Access requires 30. The effect this has on the PABX market is clear – if 30 channels are needed to make full use of primary access, then many of the smaller systems simply will not have the capacity. And here, the tariff structure comes into play. For instance, users will be billed not only for connection cost, traffic and services, but also the number of channels used. So in theory a situation could arise where a small firm, which only actually wants to use 10 channels, nevertheless must pay for the full 30.
