Plan for anywhere access by 2005

By 2005, over 40% of the $1.3 trillion (£0.9 trillion) global market for Internet services will stem from pervasive computing,...

By 2005, over 40% of the $1.3 trillion (£0.9 trillion) global market for Internet services will stem from pervasive computing, which hinges on the "anytime, anyplace, anywhere" principle, according to a report by analyst company Ovum

The authors of the report have identified a switch from current service delivery mechanisms, based around the "stovepipe" principle, which distinguishes the delivery channel and device used, to a more amorphous, service-based model.

The idea is that future services will know who the user is and deliver a personalised experience, no matter how the user accesses the service. This contrasts with today's model where users have to modify their behaviour when accessing services from different devices using different channels.

It is here that the concept of context comes in. Context in a pervasive computing environment enables the Internet service, or collection of services, to understand not only who is accessing them but where they are accessing the services from and what the user is accessing them on.

This was a key theme in Sun Microsystems' announcement of its Sun One Web services strategy last month.

Ovum has identified three main waves of technological development to support this new trend: nCompanies will create new interfaces to existing applications.

This is already happening with the introduction of Wap-based access to legacy software applications nNew, multichannel applications will be created specifically to take advantage of pervasive computing concepts nOld applications will be re-engineered to work within the new multichannel environment.

The main issues for companies wanting to take advantage of pervasive computing are management of the software infrastructure and an understanding of the delivery channels.

For organisations that have been Internet-enabling applications, migrating their software architectures to a pervasive computing model will not be as difficult as for those who have yet to make the leap from mainframe or client/server environments.

According to Ovum, major building blocks in pervasive computing software architectures will include the client operating system; service brokering, security and identity management; and quality of service optimisation.

This last issue could be the most important, because maintaining quality of service over indiscriminate network links will require the co-operation of different service providers to maintain performance on an end-to-end basis.

The establishment of quality of service mechanisms will be assisted by consolidation in the Internet backbone and by the general adoption of IP across different delivery channels - namely, the telecommunications sector, IT industry and broadcast media.

As service providers with experience in one delivery channel buy into the pervasive computing movement, it will be necessary for them to investigate the other delivery channels. And they will need to learn how to work with additional partners.

Device manufacturers will face the same challenge. "For example, automobile and home appliance manufacturers are now working with software companies, network operators and content providers to enable the delivery of applications and services to vehicles and domestic appliances," said the Ovum report.

The pervasive computing movement is still very immature, and most companies have only just come to grips with the concept of Internet-enabled e-commerce applications.

Producing services that are even more flexible and accessible via other channels will be a new challenge for them. Hopefully, service providers and software development companies will be able to make the transition an easier one.

Read more on Business applications