In-Depth

Emerging technology: Wireless computing

Wireless computing has often been discussed in roundups of emerging technologies. It seems like an obviously good thing, and more obvious since the world has been invaded by cellular phones. But the steps forward for wireless computing have been slow.

»Intro
»Grid computing
»Rich clients
»MDA
»Wireless computing
»Agents, et al
»Burke's laws
Maybe we are just looking in the wrong place.

Wireless fleets of taxis go back to before the Swing Era. Organizations with service fleets are ready to move to wireless computing today, although they may have to build hybrid systems that allow varied means of communication, as wireless network infrastructures still have some holes.

In relation to some other emerging technologies, a vast lot of vendors are interested in this area. That has not proved altogether a positive circumstance. Companies on this road include chip makers, phone makers, phone service companies, software companies and more. There are many competing protocols and, unlike the World Wide Web, there has been no sudden rush to support a single approach.

XML and Web services may play a constructive role. Some wireless project managers have been stumped by the requirement to build a whole new application base infrastructure. An approach that allows the development team to focus on new I/O paradigms without massive computer system rewrites will show appeal.

Among the many targets of Microsoft's Visual Studio .NET environment, and a first test case of Microsoft Web services, is wireless computing. "People want a single, standard integration layer through Web services," said Ed Kaim, product manager, Microsoft. The company has produced a lightweight app framework (the .NET Compact Framework) that lives on small devices. Developers can take Visual Basic skills and begin work in the wireless environment. That means cost savings, which do not hurt emerging technology's chances.

Xerox Global Services took the .NET wireless leap with a recent Field Service Asset Management application. Three developers worked two months to build the app. They created a Pocket PC app in C# using the .NET Compact Framework, with SQL Server as the database and SQL Server 2000 Windows CE Edition as a mobile device store.

The Xerox developers had already gained experience in C#, and were able to port code to the new application.

"For us, the application was a way of growing operating efficiency," said Bob Gerardy, manager, Xerox Office Service Operations. "We looked for a way to reduce manual input, create reports and to provide real-time information on the devices." If a network printer or copier cannot be fixed during a service visit, the system automatically generates a follow-up job ticket, according to Gerardy.

The device is architected so that, if an 802.11 LAN is not available, the device can synchronize "off-air" when it is placed back in its home cradle. "We did our homework learning the various hardware platforms and picked the Pocket PC," Gerardy said.

Wireless coda: Wireless computing is far from monolithic. A surfeit of big vendors is not a blessing. Legions of vendors have not seen reason to pick a single side, as they did with the World Wide Web. Also, a troubled economy has put the age-old search for "the killer app" pretty much on hold. Services that ask the subscriber for another $30 to $50 per month are sunk for the moment. Yet wherever manual input still occurs, wherever premier service wins or saves deals, wireless will get a hearing. The development effort must usually leapfrog off existing systems and the best candidates may be surprises -- a bar code system, for example.

Prev | Next

Intro | Grid computing | GUI | MDA | Wireless | Agents etc | Laws