In-Depth

Mobile Computing: Are We There Yet?

• By Art Dearing
• 05/01/2001

Almost overnight it seems industry journals are being dominated with articles heralding handheld computing as the next great revolution in the industry. We read of fantastic consumer capabilities, such as the ability to comparative price shop across the Internet by simply scanning the UPC code of an item of interest. Further, the proof of concept demonstration done in Finland last year of purchasing a soda using a Web-enabled mobile phone has already become a commercial – creating the idea that handheld computing is mature and ready for mainstream adoption.

While handheld computing has made great strides, (remember the Newton?) there are still obstacles to overcome before these devices enjoy widespread adoption as part of a corporate computing infrastructure. Bandwidth, security and connection infrastructure are some of the many problems facing companies considering handheld computing solutions. Perhaps, one of the greatest challenges is recognizing these devices for their inherent strengths and weaknesses and correctly utilizing them. This is not the first time that the computing industry has successfully dealt with this type of adoption problem, after a false start or two.

A good analogy of this is what happened as PCs began to be adopted into the corporate computing landscape. The first PCs to land on desktops during this time were quickly given terminal emulation software, allowing them to work within the "mini" and mainframe corporate computing environment, but effectively eliminating the architectural advantages of the PC. Later, the advent of the PC LAN and client server technology created an opportunity to exploit the natural advantage of the PC, including graphical interfaces, local storage and local processing.

Handheld and wireless computing is facing exactly the same adoption problem. Handheld computing devices are fundamentally different from any other computing platform in widespread use, and these differences must be recognized and exploited before these devices will be successful as an extension to corporate computing. In other words, these devices are not just smaller, less powerful versions of a desktop PC, they are different, with different strengths and weaknesses. A Web-enabled phone with its tiny display will never make a good general-purpose Web-browsing tool, even if the user is stuck riding a train for hours. And, the thought of using a 12-key touch pad to enter anything but the most trivial amount of data is almost unbearable. There is, however, extremely valuable information that can be delivered through the wireless Web that is perfect for devices of this size. And, just as with their early adoption of the PC, there will be some industry experimentation required to determine what that valuable data is for Web-enabled phones and other handheld computing devices.

One of the early general use trends of handheld computers was "Web Clipping," or resizing and formatting Web pages for display on handhelds. While offering some utility, this idea still has some limitations. Many pages do not translate well to these devices. I remember some months ago excitedly loading a page from a popular map site to my PDA, only to be disappointed later when I couldn’t distinguish anything on the smaller, lower resolution screen. Conversely, stock quotes and less graphical information translate quite well. The yet-to-be-realized extension of Web clipping is fully interactive Web browsing, using a PDA or Web-enabled phone as an ultra-thin browser, creating the "Web everywhere" experience.

Limitations with network reliability, security and other concerns put this idea somewhere out in the "coming soon to a theater near you" category. One simple example of current limitations – imagine having your top salesperson sitting in your largest customer’s main office, deep in the center of a large office building, struggling to answer account questions properly because the building blocks the signal your company uses to connect PDAs to critical corporate data. If your salesperson has to say, "Let me step outside to get you that quote," it suddenly kills all the advantages of supplying a sales force with connected PDAs.

Web-connected handheld computers hold great promise to be sure, but there will be some experimentation required to get things right even after infrastructure issues are resolved. Location commerce or proximity advertising is an interesting extension of general Web access on handhelds. This idea suggests that as a person moves through a cell site, their approximate location can be tracked and information (read: advertisements) from local vendors can be automatically forwarded to that person’s Web-enabled phone. In its best form, this will be a wonderful way of providing timely information from your favorite restaurant or store. In its worst form we now have a new way to create and distribute "spam on the go."

In addition, the idea of using a handheld strictly as a Web interface needs to be balanced with the native capabilities of the device itself. Remember, Palm devices gained popularity long before they could be connected to the Web. This seems to reinforce the notion that there is intrinsic value to this radically different computing format. Ignoring these basic strengths by using these devices just to connect to the Web is missing the point. History is a good case in point. Soon after Netscape made its appearance and people started to understand the impact the Web would have, computer articles began to appear, stating things like "Desktop applications and OSes are dead!" and "The browser is the window to all applications and all data."

Despite these warnings, Microsoft survived the browser onslaught just fine, federal court issues aside. The inherent strengths of a PC, local storage and local processing still require applications and an operating system. And, likewise, handheld computers and PDAs, have inherent strengths. Their fast, easy access to critical data have made these devices popular, even as notebook computers have continued to get smaller, faster and cheaper. Accessing data only through the wireless Web cannot guarantee fast, easy access to important data. As a result, locally stored information will continue to play an important role as this platform continues to develop.

Yes, connected handheld computing will be big, despite the current limitations. How do we know? First, the raw market size potential. Industry figures vary, but most experts feel that the current installed base of mobile phones, messaging devices and PDAs is several times bigger than the current installed base of PCs. Second, nearly everybody who’s anybody, including a whole lot of nobodies, has announced products, services and strategies for mobile computing.

Big name players, like IBM and Microsoft, have announced and released strategies and products for this hot new market. Most major database companies have released (or talked about releasing) ultra-thin clients designed for mobile computing devices. Many of the dotcoms have added or repositioned toward mobile computing, and there are plenty of new upstarts with names that are barely recognizable to further add to the confusion. Most of these companies are focused on building infrastructure – the connection pieces necessary to connect these handhelds to corporate data. Make no mistake, these early offerings fall very much into the "software version 1.0" category, generally requiring significant customization to get up and running. What are the best options? Current thinking seems to be that it’s too tough for mere mortals to add mobile computing capabilities to a corporate computing environment. As a result, the ASP model has been introduced to this market segment to allow companies with mobile computing technology to offer those technologies as services, complete with the custom programming that is currently required to fill in the gaps. Aether, Palm and Avantgo are some of the industry leaders that have adjusted their strategies of selling software licenses for mobile enterprise solutions to offering those same technologies as services.

Two particular gotchas in deploying a mobile computing solution are 1) correctly developing applications that run on the mobile device and 2) enabling access to the corporate data that is necessary to run those applications. The first potential difficulty, correctly developing applications that run on a mobile device, is a perception issue. Excellent tools exist for the creation of applications on the current industry-leading Palm computing platform. The potential difficulty lies in correctly understanding what types of applications work well in this environment and what types do not. As previously mentioned, this is not just a smaller, slower PC. Devices, like the Palm PDA or other PDAs based on the same technology, have the same general strengths.

Fast access to important data. A Palm device does not require the lengthy boot cycle common to most desktop or notebook PCs. As a result, at the touch of a button, addresses, phone numbers or other reference information is instantly available, making these devices far more usable for quick information checks. This feature also defines user expectation as to how applications should behave on these devices. Delays, of even a few minutes while data is wirelessly retrieved from a server somewhere, create the impression that the device is too slow and cumbersome to use.

Simple interface. This platform strength is in contrast to the feature-rich environment of a typical PC operating system where we left click for this, right click for that, double click, single click or drag and drop. The screens for PDAs are smaller than PCs, contain less information and house applications that when successful are easy to understand and navigate. To a point, PC applications do not translate well in this environment, and we should be ever grateful that this is true!

The second gotcha, enabling access to corporate data, is a result of the early (or nonexistent) state of the tools necessary to provide this function for handheld devices. The explosion of the Web and glut of content available has created a genuine content management concern for many IT managers that reaches back into the core IT infrastructure. Adding handheld computing devices to the corporate computing infrastructure only compounds this problem. Underestimating the amount of the data, difficulty in finding and retrieving content from scattered databases and efficiently propagating data to handheld devices are just some of the challenges that must be overcome to effectively connect these devices.

On Palm or Palm-compatible devices, this process involves creating something known as a conduit. Conduits provide several functions in moving data to and from these types of devices. A conduit defines exactly what data is being moved, from what source and handles conflict resolution during an update process. Almost without exception, current conduits must be built from scratch. To add further complexity to an already difficult problem, multiple data sources require multiple conduits. A company that is 70 percent Web-enabled must develop separate conduits to back-end databases and front-end, Web-enabled data. In a complex corporate computing environment where important data must be pulled from many different sources, the process of creating and maintaining conduits is daunting and adds time and complication to deploying a mobile computing solution. An additional problem that currently exists when considering conduit development is the relatively simple technical capability of the conduits themselves. Conduit technology lacks more complex data handling capabilities, such as change detection within a defined data set. Without change detection capabilities an entire data set must be replicated to a handheld device when only a single data entry has change. This puts further strain on an already tenuous wireless connection and forces design constraints when developing mobile applications.

To add to this particular headache, Palm developers are very hard to find. Once again, current wisdom dictates that this work be outsourced to a consulting firm or included as part of a service agreement from an ASP. The difficulty lies in maintaining the conduit as data requirements change. Companies considering wireless solutions will almost certainly need to experiment in order to hit upon that killer mobile application. Once again, significant development work is required, making this process time-consuming and tedious. Help does appear to be on the horizon as new tools, specifically designed to help in the creation of conduits, are brought to the market.

Lest the current mobile computing landscape start to look to disparaging, please remember that for all the articles, TV commercials and companies claiming to have solutions this is still a very early market. Analysts are just now talking about major inflection points and hockey-stick growth in the months to come. There are shining examples of companies that have successfully deployed wireless solutions. Schwab’s PowerBroker service, for example, brings the power of the Internet and online investing to a person’s fingertips by coupling a Web-enabled PDA or cell phone with a wireless Internet access service. The company has been working on its wireless Web service for some time, positioning it as a supplement to its existing browser-based online service. Schwab has partnered with Aether Systems Inc. to provide the network connectivity and the translation software that makes all of the existing Web content available to wireless users. Other companies from other industries are also successfully experimenting with wireless mobile computing solutions. We think nothing of the fact that Federal Express has been very successful using custom mobile computing devices to scan and track package delivery for some time now.

So, perhaps the ideal of using a Web-enabled phone as a shopping appliance (purchasing that cola without money) is not that far-fetched after all. Solutions will get better, standards will emerge and the ultimate promise of mobile computing will be realized, allowing companies the opportunity to get even closer to their customers – in fact, right into their pockets.