Contributed editorial appearing in
Scientific Computing & Instrumentation 19:5, April 2002, pg. 16.
The IEEE debuted a new peer-reviewed journal this year on the topic of Pervasive, a.k.a. “Ubiquitous,” Computing. The inaugural issue, “Reaching for Weiser’s Vision,” is a tribute to Dr. Mark D. Weiser, who, in 1991 published a groundbreaking article titled “The Computer for the 21st Century.” The vision described by Dr. Weiser was not that of a large, all-powerful behemoth of information and wisdom commanding the attention of armies of technicians and megawatts of power, but exactly the opposite — a collection of tiny, independent devices that “calmly” perform their duty of serving mankind while maintaining such a low-profile existence that they “disappear” into the fabric of our surroundings.Mark Weiser joined the Computer Science Laboratory at Xerox Palo Alto Research Center (PARC) in the late 1980s and rose to Chief Technologist at PARC before his short battle with cancer in 1999. He initiated PARC’s Ubiquitous Computing Program and championed the idea of technological advances leading toward “calm technology.” Dr. Weiser and his colleagues coined three eras of computing technology. The sharing of scarce computational resources exemplifies the initial “Mainframe Era.” Scarcity relegates the operation of the computing resources to a select few tapped to maintain the systems and parcel out slices of access time to those duly certified and deemed deserving. This was followed by the “Personal Computer (PC) Era,” wherein each user could interact with their own private computer. Analogous to the automobile, individuals can own and operate multiple PCs, but typically only one at a time. The widespread use of the Internet and Distributed Computing is facilitating a transition into the next era dubbed “Ubiquitous Computing” (UC).
In the UC era, individuals may unknowingly utilize hundreds of computers while going about their daily lives. Weiser likens the use of computers in the UC era to our commonplace use of electronics today. While commanding our attention when first introduced, it is easy to take the machines that brew coffee, toast bread, refrigerate cream, wash cups, open garage doors, and start automobile engines for granted while we are preoccupied with planning our work day. These electronic marvels have disappeared from the center of our attention and moved into the periphery. We interact with them calmly and effortlessly until they break down or hail our attention.
Data acquisition and measurement systems have evolved in a similar fashion. The days of collecting samples, carefully preserving them, and shipping them off to a remotely located analysis laboratory, whether leagues or buildings away, are thankfully disappearing. As analysis instrumentation continues to miniaturize, we are gaining the ability to measure things where they are found. Increasingly, measurements are being made in the same laboratory or production floor that is producing the samples, thus moving us into the “Personal Instrument Era.”
Distributed Data Acquisition (DDAQ) and Laboratory Information Management Systems (LIMS) are permitting us to transition into the “Pervasive” or “Ubiquitous Instrumentation” (UI) Era. Also gone are the weeks of training on a new “instrument,” requiring both talent and experience for proper operation, not unlike a musical instrument. Onboard analysis routines free the technician to focus on more important matters. The challenges of the UI era are not limited to development of small, remote sensors. Network communication architecture and models used to operate a large collection of active sensors in concert pose even greater obstacles. This pervasive era, being the inverse of the mainframe era, must deal with the time-sharing logic required to cope with many computers and instruments wishing to command the attention of the scarce commodity known as the “user.”
In the UI era, sensors and instrumentation must be infused with a level of intelligence needed to function autonomously in a default mode of “seen and not heard.” While operating in a UI community, instruments should perform calibrations automatically and routine tasks such as the refilling of solvent bottles should be completed without user intervention. Autosamplers must execute their duties tirelessly and only alert the user when unexpected or problematic results are obtained. Let us hope this new journal of Pervasive Computing continues to incubate visionaries like Dr. Weiser.
