When the semiconductor giant Intel installed a WiMax wireless network at the UK's Science Museum's airfield campus in the Wiltshire countryside, Sally Pettipher, the organisation's head of development, was so delighted that she felt like dancing on the runway in the moonlight.

For the first time, staff have permanent, high-speed, wireless Internet access throughout the aircraft hangars that are used to house thousands of stored exhibits. Before the combination of WiMax backbones and WiFi hotspots was installed, staff were driving information between the hangars and offices in cars. Now their jobs have become far easier.

Over the coming years, Pettipher's delight is likely to be shared by tens or perhaps hundreds of millions of others as high-speed, mobile Internet access reaches areas that have hitherto been inaccessible, primarily because of cost. Throughout large buildings, and across campuses and entire metropolitan areas, always-on wireless Internet connections will become routinely available.

However, the opportunity is complicated by uncertain economics and difficult technical choices. Hype, confusing operator strategies and uncertain regulatory issues are clouding the picture.

Two wireless technologies will play a critical role. The high bandwidth technology adopted by mobile service operators, 3G UMTS, is widely understood but still in the early stages of deployment after years in development. The other – WiMax – is only now emerging.

If WiMax fulfils the forecasts of its evangelists at Intel, it will drive down wireless Internet prices, attract millions of users and enable a new generation of operators to offer low cost alternatives to 3G mobile services and wired connections. WiMax is not only capable of supporting higher speeds than 3G (see box), but networks can easily extend the reach of WiFi networks from metres to kilometres. Critically, it is based on open standards, making commoditisation inevitable.

This could give WiMax an important price advantage over both 3G and fixed line DSL services. Over the next two years, to 2007, fixed (stationary) WiMax technology will mature and fall in price to compete with DSL services. Meanwhile, a new mobile version will become available, included in Intel's Centrino chip technology that is used in most new laptops.

That development will be critical: the availability of low-cost Centrino chips in 2003 was a "turning point" for take-up of WiFi technology, says Niall Murphy, CTO of WiFi operator The Cloud.

In spite of this, few predict that WiMax will have a huge effect on the telecoms operators. Nor is it widely expected to drive down prices dramatically, or push up connectivity speeds. Ian Keene, an analyst with IT advisors Gartner, says WiMax could conceivably be "just another niche technology that will hang around for a while and then disappear".

With the technology proven, the backing of giants such as Intel assured, and prices certain to fall, why the uncertainty?

There are at least four factors that will impede rapid take up. The first of these is availability: While WiMax chips will be available in 2007, they will still be rare in PCs until 2008. That suggests high volume take-up will not begin until 2009.

The second is price and ease of installation: At present, WiMax equipment costs around e200 per customer device, and expensive external aerials must be bolted onto buildings. In comparison, WiFi chips cost around e5 per user and are easy to set up. That gap will close, but not soon.

By then, a third factor comes in. 3G service operators should have millions of customers and may well offer integrated WiFi/3G roaming. At this point, they are likely to offer tactical use of WiMax services to those that need high bandwidth roaming. Operators are already planning a new 3G service, HSDPA (high speed downlink packet access), that will offer 3.4 to 14.4 megabits per second.

A fourth factor is also causing uncertainty. WiFi networks are not regulated, but WiMax networks may well be. In the UK, Ofcom is reviewing radio spectrum allocation. Some networks will remain unlicensed, but the more powerful and better networks will be based on paid-for licenses. That will push up costs.

Taking all these together, there is a consensus emerging about WiMax that hardly justifies much of the excitement: it will become just one of many technologies used in an always on, mobile world. Taken alone, it won't trigger a revolution.