Two projects currently underway at the University of Pittsburgh’s Department of Electrical and Computer Engineering are aimed at helping companies customize passive or active RFID chips for specific applications.
Led by Professor Marlin Mickle, university researchers are developing two design programs. One program would generate code that would then be written to a general-purpose (off-the-shelf) chip used in an RFID tag. This code could add data storage capacity to the chip, for example, or add functionality related to how the chip in an active tag communicates with an attached sensor, and how it transmits that data to an RFID interrogator, says Mickle. The other program being developed would be used to design a tag’s chip from the ground up, thereby building any special functionality, as well as the required data storage capacity, right into the chip.
Using these programs, companies could either retrofit off-the-shelf RFID chips or custom-design chipsto match their own needs. Mickle says this would eliminate the cost and time associated with hiring an engineer, or staff of engineers, to design a chip used in a tag to perform specific functions. The production of the chips and/or tags, however, would need to be outsourced.
In developing the method of retrofitting a general-purpose chip, Mickle says, the university partnered with Wexford, Pa.-based semiconductor design firm ADCUS (a subsidiary of South Korea’s Advanced Digital Chips). The university is in the final testing stages of the program, which generates a set of code that is written to the tag’s chip through RF energy, similar to the manner in which an EPC is encoded to a tag.
ADCUS optioned the intellectual property behind two patents that the university has filed as a result of this research project. The patents address how to write the software program needed to customize the functionality of a chip used in an RFID tag. Once the project is complete, ADCUS will have 90 days to review the results of the project and negotiate the terms around licensing the IP. ADCUS hopes to develop a commercial product based on the technology; it does not manufacture RFID tags, but the company’s CEO, John Kang, says his firm is very interested in developing innovative approaches to active RFID.
“The University of Pittsburgh was an ideal partner for us,” says Kang.
Mickle and his team began developing the second program—using software to design custom chips—in May, with university funding (it is currently looking for commercial partners). The vision behind this project is that rather than taking an existing chip and optimizing it for an RFID application, a company would purchase software to design a layout for the silicon chip from the ground up. Mickle and his team are developing an easy-to-use interface for the design software, which he claims will function like a CAD (computer-aided design) program. The software will reportedly incorporate the requirements of air-interface protocols so that the finished chip design will satisfy international tag standards.
According to Mickle, his approach, which the university calls Primitive to Silicon, will be especially beneficial when designing chips for passive tags, which do not contain batteries and must consume as little power as possible from the interrogator. With the Primitive to Silicon software, a user could design passive chips with a power requirement lower than that of general purpose chips.
Mickle says the Primitive to Silicon method of designing customized tags would also be more cost-efficient than customizing general-purpose chips if the user requires a large quantity of tags—1 million or more. Once designed with the Primitive to Silicon software, the prototype chip would be made and tested through an outsourced silicon foundry. After any required changes were made to the prototype chip, its production and the tag-assembly process (during which the chip is attached to an antenna and placed on a substrate) would then be outsourced.
The university will receive royalties if the licensed IP behind either of these design programs is used in products or services sold by ADCUS or another licensee. However, Mickle says that by developing new approaches to RFID tag design, the university is also remaining at the forefront of the technology. Plus, he adds, the university would also be helping companies get a foothold in the RFID market by offering tag design or customization services based on these design programs. “It’s difficult for small companies to penetrate the RFID tag manufacturing market because they don’t necessarily understand the technology. These initiatives could allow them to take their knowledge of a particular market and actually develop a customized tag,” he says.
This is not the first RFID project Mickle has initiated at the university. In 2002, he unveiled a design for a Product Emitting Numbering Identification (PENI) passive tag (see A New Approach to RFID). Rather than using one antenna and the conventional backscatter method of transmitting data between a tag and reader, the PENI design uses two antennas on a tag—one to transmit data, and one to harvest power from the electromagnetic field generated by the interrogator. Tags that use backscatter typically reflect the radio frequency energy emitted from an interrogator as a modulated wave. That wave pushes through energy coming from the interrogator, which is a less energy-efficient process than the PENI method. According to Mickle, development work on the PENI has been put on hold because it is considered a disruptive approach that the RFID infrastructure is not ready to accommodate at this time.