To coin a phrase from a popular book, the next "new new thing" promises to be high-bandwidth, ubiquitous wireless access to the Internet. The uses of this bandwidth will be as diverse and as impossible to predict as use of the Web is today. They will include navigationally aware cell phones; intelligent, interactive, content-rich vending machines; fully connected roaming professionals conducting mobile video conferences anywhere their palm device batteries can take them; and a million other scenarios that will spring to the minds of inventive people with the vision and the drive to make them a reality. While many future uses are but ideas on bar napkins, there are many high- bandwidth wireless applications in production today in Japan and Europe, and notable numbers under development in the U.S.

Let's examine a typical scenario: an on-call doctor accessing a health system's federated business intelligence (BI) system. The on-call doctor receives a page via her wireless device. Using the same device, she calls the hospital and learns that another doctor's patient, whom she has never examined, is experiencing difficulties in the recovery suite after minor surgery. The floor nurses need to know if they should increase the current pain medication or switch to another type that could be more effective. The on-call doctor turns to her wireless device and calls up the patient's computerized record to review the current diagnosis, the x-ray and CT images, the surgical procedure, the anesthesia used, the patient's past medical history, the current prescription drug mix, the list of available alternative pain medications, the automatically generated list of alert flags for mixes of medications that could be problematic for this patient, the patient's allergies, the real-time data feed from the heart and blood pressure monitors attached to the patient, the standard medical treatment path (best practices) for this diagnosis and the available resources related to this condition in the internal and Internet-based knowledge repository of medical journals.

In minutes, the on-call doctor reviews the available information, changes the prescribed medication to one that is less likely to conflict with the blood pressure medication the patient is receiving, sets a reminder flag to check in on the patient's condition in six hours, updates the patient's chart, sends a notice to the attending physician and returns to her tee shot on the seventh hole.

In this brief interaction, the on-call physician accessed data from real-time clinical system feeds, clinical OLTP systems, OLTP vendor-supplied turnkey data warehouses and data marts, custom data warehouses and data marts, third-party data warehouses and external Internet data sources. All of this information was presented in a seamless, coordinated interface via the enterprise information portal structured for use in the wireless environment. In addition, each data source was optimized for this environment by providing a return data stream that would not overwhelm the relatively limited delivery bandwidth, processing capability or the display size, resolution and format of the mobile device.

This sample application of wireless device access to the BI infrastructure in a closed-loop manner is not a fanciful dream, but a very practical and focused goal of several of our health system clients. Our commercial clients are working just as hard to provide wireless access to their BI systems for remote users, executives and mobile sales professionals.

In this example, as in all of our work in wireless warehousing, the primary system design trade off is between speed, scope and summary.

Wireless access is marked by the demands of extremely quick turnaround of data requests. The nature of wireless access is time constrained; users do not have the luxury of waiting four minutes, 40 minutes or four hours for an answer. Thus, speed is paramount.

If we are designing a system to provide detailed, focused information to a wireless user, we must either partition the data into relatively small segments that reflect users' access needs (i.e., by recently admitted patients in this scenario) or provide for extremely quick performance by maintaining a highly indexed environment along the primary "where clause" dimensions of the data set. In our example, it is very challenging to provide integrated high-speed response to a patient's medical history, current chart and procedure detail across the variety of required source systems, to say nothing of the real-time data feeds.

If we limit the access to highly aggregated information in order to meet the performance demands, then we risk not providing enough information to serve the needs of the users. In our on-call physician scenario, an aggregation of the incidence rate of postoperative infections or drug interaction adverse outcomes would be of little use. Detailed information is required.

Wireless device access to the BI infrastructure requires a well-designed, federated BI architecture to integrate data across the variety of real-time, turnkey, custom and third-party data warehouse/data mart systems found in the typical enterprise. It also requires careful tradeoffs between speed, scope and summary in the design of the data sets themselves.

The next great wave of BI utilization promises to be Internet-enabled, wireless devices. It is best to understand the unique challenges of this environment today, rather than tomorrow.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access