J2ME Makes Its Mobile Mark

When the idea of bringing Java-equipped handsets to the United States was first proposed more than three years ago, some observers expressed concern that it would be received here with the same chill with which WAP was received. Could anything useful really be done with small displays and fiddly keypads, not to mention the small memory size and processors that are relatively puny compared to PDAs?

Those doubts are being to be laid to rest as the first wave of Java 2, Micro Edition applications prove their value--not merely for games and other consumer applications, but for real enterprise applications that deliver measurable improvements in productivity, accuracy, and customer satisfaction. Furthermore, by using off-the-shelf mobile phones, conventional wireless phone networks, and the widely supported J2ME standard for development, these handsets are delivering enterprise value with previously unseen levels of ease and economy.

One such example is the application built by AirClic Inc. for Boston's public schools. The school district's attendance officers spend much of their time on the streets looking for truants. They refer to a list of students to check where a particular student should be and whom to contact. Before J2ME came along, the list was a computer printout carried by the officer. Like most directories, the list was out of date almost as soon as it was printed. What's more, the 63,000-name list was the size of a phone book, making it cumbersome and the process inefficient.

AirClic teamed with Nextel to replace the printout with a Motorola J2ME phone on the Nextel network, which the attendance officers have been using in a trial since May. Officers tap student names into the J2ME application running on the phone and connect to Boston Public Schools' back-end systems via AirClic's Mobile Information Platform middleware. The middle tier returns identifying information including age, address, attendance record, and parents' phone numbers. Planned enhancements to the system would show a student's bus routes and medical information such as allergies. With access to the student's schedule, the officer can even see which class a student is cutting.

AirClic also offers a small, lightweight bar-code scanner that attaches to the phone. As the school district rolls out bar-coded school ID cards, officers will be able to use the scanners to identify a student more quickly. The usefulness of the scanner goes beyond conventional uses such as identifying students (or, in other applications, warehouse inventory or school supplies). AirClic also has turned the scanner into a control mechanism: Users can scan bar-code sheets with instructions like "identify student" and bypass menus and other keypad input altogether. Several other school systems have expressed interest in the service, as well as in other uses of the same technology. One example is tracking students on and off buses (among other things, federal subsidy dollars depend on reliably auditing bus use).

In another example, mPortal Inc. has worked with Xerox to create a field-service application designed to be more convenient for engineers to use, more connected for more-prompt data reporting and more effective scheduling, and considerably more cost-effective to deploy.

The engineers, servicing Xerox equipment, had been using laptops to load daily call schedules. MPortal created a J2ME application, again using Motorola phones on the Nextel network. (Nextel was the first U.S. carrier to offer J2ME phones so it's no surprise to find most of the early success stories among its business customers). The application lets service engineers work on- or offline, synchronizing to a central server via mPortal's Enterprise Everywhere platform as network coverage permits. Consequently, they can close more work orders a day and work more dynamically, boosting productivity and helping more customers.

The key to the Xerox project was the emphasis mPortal and Xerox placed on the front-end stages, analyzing and understanding the business case as well as the work tasks of the engineers. The two companies estimated the return on investment before even attempting to design an application.

Xerox spent considerable time "riding along" with engineers to refine the application and ensure the result would cover all business processes. Once the requirements were thoroughly understood, focus moved to the user interface, both in terms of user performance and user acceptance. It was important to persuade users that the task could reasonably be done with a small screen and keypad instead of the familiar laptop.

Designers focused on what the core activities of the engineers on site were, and then studied how to make those activities as simple as possible. Some of the techniques used included pick lists, numeric codes to minimize data entry, and "jump" screens to simplify navigation as the users' familiarity with the application grew. The team went through three iterations of development and user feedback once the interface was finalized--a not-uncommon experience for those teams that care about the usability and effectiveness of an application on a small device.

These examples are two among many, and more are continuing to roll out. The best-candidate applications for simple, cheap J2ME handsets are those where a small, simple interface can exploit heavyweight processing on the back end. Success requires that designers emphasize usability in the broadest sense, with a major focus on requirements gathering and task analysis. With the limited interactivity of a handset, the application has to match extremely well to the needs and activities of the user. Even with the addition of bar-code scanners to input both actions and data and so bypass some of the limitations of a keypad, it's still essential to understand how users work in the field to provide them with the right actions and shortcuts.

In the right application, a relatively modest incremental investment in hardware and software can return significant gains in productivity, accuracy, and responsiveness. If you've set aside projects in the past because of the cost of equipment or deployment, or the difficulty of wireless connectivity, reconsider those projects in the light of the economics of J2ME handsets.

Carl Zetie is VP of research at Giga Information Group.