The most technically challenging aspect of the Memory Spot project, explains Taub, was to integrate all of the components – an antenna, a modem, memory, and a processor – onto a chip less than a millimeter wide. All the components are placed in a specific configuration to keep the antenna from interfering with circuitry in the rest of the device.
Part of the reason this high capacity is possible, explains Taub, is the wireless frequency used to transmit data to and from a tag via the Memory Spot reader. The chip’s antenna and reader operate at the 2.45 gigahertz frequency band, which is used in Wi-Fi, while most RFID tags use 13.56 megahertz. By operating in Wi-Fi range, significantly more bandwidth is available to send data from chip to reader, resulting in transmission rates of 10 megabits per second, much faster than traditional RFID. Data transfer speed is a crucial feature because “people wouldn’t want to wait around for a couple of seconds while the reader is reading the chip,” he says.
Additionally, how data is transmitted is an important consideration, says John Waters, a researcher at HP Labs in Bristol, U.K. who worked on the project. “We have specifically designed a scheme that…requires a minimum amount of circuitry in the chip to function,” he says.
The HP announcement is another step in the evolution of wireless, ubiquitous computing, says Lionel Lavallee, senior RFID solution architect at Intel. Looking back at the types of RFID available a few years ago, the devices were fairly distinct, he says, and could be broken into a few categories: passive RFIDs that use power from the reader and hold only a small amount of data; active RFIDs that have more computing power and memory, but also use batteries that have a limited lifetime; and battery-powered sensors that collect information from the environment. “Now you’re seeing a continuum of all the technologies bleeding together,” he says. “You get this cross pollination that grabs the best bits of both worlds.”
Lavallee predicts that RFID will finally start to become widely adopted in businesses in the next 12 months. “Right now we’re at a tipping point,” he says. Standard RFID technology will start to become commonplace, showing up in hospitals to track patients and treatments, in pharmaceuticals to prevent counterfeiting, and increasingly on products to replace bar codes.
And Memory Spot could add another set of applications to wireless chips because of its much greater memory capacity and battery-free longevity. Still, HP’s chip will not completely replace traditional RFID, suggests Taub, since it’s more expensive than conventional RFID tags, which can cost less than 10 cents each. Additionally, not all applications – for example, authenticating a shipment of flip-flops for Wal-Mart – would need so much memory.
Nonetheless, the chip’s relatively large storage capacity makes it an ideal candidate for some wireless technology and applications. While it was conceived as a way to add audio to photographs, says Taub, he can envision attaching the chip to a postcard, to send pictures of vacations to family and friends, or being used in textbooks to create media-rich environments.
At this early stage, however, it’s not clear what applications Memory Spot will be best suited for, if it does make it out of the lab. “It’s hard to predict a killer app until it’s released in the marketplace,” says Rajit Gadh, professor of mechanical and aerospace engineering at UCLA and a specialist in RFID technology. But he’s encouraged by the capabilities of the HP chip and expects that one of its most exciting applications will be for storing and sharing digital media. “I think this is a very positive development for the field of RFID,” he says, “with the possibility of creating new markets, such as media streaming content over passive RFID.”