This was the impetus for the 1451.4 standard, developed by the IEEE in conjunction with companies such as Aeptec Microsystems, Bruel & Kjaer, Crossbow Technology, and my company, National Instruments Corp., as well as the National Institute of Standards and Technology and the U.S. Air Force, among others. The standard is built around the concept of the Transducer Electronic Data Sheet, or TEDS.

A TEDS describes the sensor's electrical interface requirements—how the data acquisition hardware needs to be configured to read the sensor properly—and tells the acquisition system how to scale analog output voltages properly into the units that correspond to the physical property being measured, such as degrees Celsius. By doing this automatically, in one swoop, a 1451.4-compliant device eliminates the possibility of human error in transcribing the data sheet.

The heart of 1451.4 is the use of a digital read-only memory (ROM) chip embedded in the analog sensor that stores the sensor's electronic data sheet, as well as information identifying the sensor—namely, its type, manufacturer, and a serial number. When hooked up to a 1451.4-enabled data acquisition system, the ROM chip transmits the TEDS to the system, in a way similar to a USB mouse or printer identifying itself to a PC after it is plugged in.

The standard, however, importantly, does not dictate where the relatively delicate ROM chip should be placed in relation to the tougher analog sensing element. The chip can be added inside the sensor housing, in the sensor connector that attaches to the data acquisition equipment, or even inside the sensor cable. This allows the analog sensor itself to still be placed in harsh environments, with the ROM chip located in the usually more benign environment at the other end of the sensor cable.

To see how powerful 1451.4 is, consider the problem faced by the state of Ohio, which has to build and maintain roadways that must endure extreme winter conditions. To research the durability of the materials used in road construction, the Ohio Research Institute for Transportation and the Environment at Ohio University, in Athens, studies the performance of pavements by embedding sensors in roadways. Analog sensors measure strain, load, and deflection in the pavement. Because of the environmental conditions and cost constraints, permanent data acquisition systems and all-digital sensors were impractical. Instead, the Ohio team decided to use a portable data acquisition system that could be transported to each of the monitoring sites and hooked up to the permanently embedded sensors.

One of the greatest challenges of the project occurs when the portable system arrives at a site. The embedded sensors have to be connected to the system to conduct a test, but with the sensors embedded in a roadway, how can the team keep track of which cable belongs to which sensor? You could write out a paper or card tag for each cable, but as researchers soon discovered, even though the ends of the cables terminated in a protective box at the side of the road, mice get in and eat the tags. You could write directly on the cables, but the rodents have a taste for PVC insulation too, and nibble at the writing. Color-coded cables might work—except the dye in the colored insulation leaches out in the salty runoff from an Ohio highway over the winter, so a red, black, and white set of cables will be white, white, and white by the time the team returns. Metal tags do work, but making them is time-consuming and costly.

To combat the problem, the Ohioans are developing a 1451.4-based system that will automatically identify which sensor is at the end of each soggy, mice-gnawed cable. They have successfully demonstrated the system in Ohio University's indoor pavement testing facility, and when the team next gets permission to tear up the highway and install new sensors in 2007, those sensors will be 1451.4-enabled.

Even if you can't install new sensors, 1451.4 may still be able to help. The standard also establishes the concept of "virtual TEDS." It allows 1451.4-enabled data acquisition systems to download correct calibration information for the billions of legacy analog sensors already in place that don't have built-in TEDS chips (just as long as someone has created a virtual TEDS for the sensor). An entire database of TEDS files for tens to millions of sensors can be stored on a disk or on a server accessible over the Internet. A unique ID number, sorted by vendor and model or serial number, identifies each TEDS.

John Deere, a major manufacturer of agricultural, construction, and forestry equipment, has a product engineering center in Waterloo, Iowa, where engineers are planning to use virtual TEDS to deal with legacy sensors. The center uses sensors for measuring such items as temperature and pressure inside new vehicle transmissions. The test center's goal is to calibrate every sensor in a central laboratory and then tag each one with a unique identification number. The calibration information will be uploaded into a database in the virtual TEDS format. Then the sensor will be dispatched to product-testing facilities, where technicians will use the ID number to download the matching calibration information. This would then be used by 1451.4-enabled software and hardware made by National Instruments, of Austin, Texas, to automatically configure the data acquisition equipment for the sensor, improving the quality of data and reducing the number of repeat test runs.