Contributed editorial appearing in
Scientific Computing & Instrumentation 21:12, November 2004, pg. 12.
"No, I just have more homework than most" was my standard reply when new acquaintances in undergraduate classes learned of my double major in the physical sciences and accused me of being "smart." I began to dread the inevitable round-robin introduction sessions so much that I often told my new classmates I was a communication major. I meant no disrespect to the COM folks. It was simply the largest major on campus and melting into the background among peers who were out to size up the competition for high grades was infinitely more comfortable than painting a bull's-eye on my chest. As an apprentice scientist being trained to discover and relay the truth, I felt compelled to rationalize my minor deception. I had plans to continue onto graduate school as an analytical chemist and reasoned the goal of that profession was to establish "communication" between analysts and properties such as temperature, pH, mass, and composition, so it wasn't too much of a stretch after all.Now that I am on the other side of the chalk, so to speak, I specifically inform my instrumentation and measurement students that we are studying the process of information flow from transmitter to receiver, i.e., communication. The physical quantities buried in the system under study are made to announce their values through their interaction with an appropriate transducer — a device that converts system quantities into electrical signals such as voltage or current. These signals carry information concerning the desired system value to the input of a model that has been calibrated to translate the transducer response into a facsimile of the desired system quantity. Like most communication systems the measurement process is susceptible to noise entering along the signal transmission lines or via incorrect calibration model parameters. To minimize noise, analog voltage and current signals are often converted to digital representation in close vicinity of the transducer and standard operating procedures minimize the misapplication of incorrect calibration parameters. This is an elegant approach, but one that generates a slew of new issues to be considered.
Specifically, once a swarm of transducers are networked to facilitate communication between themselves and supervisory command and control, what should they say? How should they say it? How fast should they say it?
Over a decade ago, in September 1993, a group of measurement folks from the National Institute of Standards and Technology (NIST) and the Institute of Electrical and Electronics Engineers (IEEE) Technical Committee on Sensor Technology of the Instrumentation and Measurement Society co-sponsored a meeting to discuss the creation of a standard interface for these emerging networked "smart sensors." Their Herculean efforts have produced several standards under the main title IEEE-1451. The initial standard, IEEE-1451.1 addresses the difficulty traditional transducer manufacturers face when connecting their transducers to the numerous and varied commercial data networks. To alleviate the development of multiple network interfaces and drivers, a common network-independent application model was developed to run on a standard interface called the Network-Capable Application Processor (NCAP), thereby answering the "how should they say it" question.
The task of answering "what should the transducers say" was championed by the IEEE-1451.2 working group. After defining the mechanics of the Transducer to Microprocessor Communication Protocol, the format and content of a Transducer Electronic Data Sheet (TEDS) was developed to be stored and read from memory attached to the transducer. The TEDS incorporates data structure information including version number and number of channels, identification information including model, serial, and revision numbers, and channel information such as range, physical unit, uncertainty, response times and clock frequency — all of the information shipped along with traditional transducers on sheets of paper that are often filed away and only looked at installation and when the transducer behaves poorly. Since the TEDS resides in user-rewritable memory, the standard also defines the storage of calibration information including structure, last calibration date-time, calibration interval, and values of model parameters. The goal of the TEDS is to simplify field installation by providing the sensor with "plug and play" capability. Legacy sensors without onboard memory are supported by the IEEE-1451.2 TEDS standard through the creation of online Internet databases of TEDS values provided by manufacturers and identified by transducer serial number.
IEEE-1451.3 answers "how fast should they say it" with standardized electrical interfaces, hot swap procedures, and time synchronization protocols while IEEE-1451.4 addresses mixed-mode transducers that communicate TEDS information digitally and then provide process values in an analog format. The most recent working group, IEEE-1451.5, is developing wireless communication protocols to keep pace with a transducer industry that only dreamed of wireless back in 1993. In addition to providing the interface for networked sensors, the IEEE-1451 standard facilitates the implementation of Condition-Based Maintenance (CBM) on the transducers themselves and the processes they monitor. Like most intricately conceived works of literature, this standard was well worth the wait.
