Good Problems

At the University of California, Los Angeles (UCLA), the RFID lab is just across the street from the medical school. This could be one reason UCLA is conducting research into RFID applications for health care. Rajit Gadh, the professor in charge of the work, and his team have developed three innovative solutions so far. Their experience says a lot about the challenges of deploying RFID in almost any setting, even when the benefits are obvious.

The first technology they introduced was an RFID system designed to help hospitals ensure that only authorized medical personnel, parents and other caregivers could move newborns from place to place. Electronic systems are available, but they tend to make it hard to move babies around. Gadh’s team developed a foolproof system that automatically matched infants and caregivers. But once the technical problems were solved, there were political challenges to manage. Hospital workers’ unions were concerned about unreasonable tracking and monitoring of their members.

Next, the UCLA team developed an RFID solution to prevent patients from getting the wrong treatment because they’d been identified incorrectly. Mistakes can be relatively benign, like putting people through unnecessary tests, or horrific, like amputating healthy limbs. The system worked, but again, there were adoption barriers. Privacy was one; in the United States, health-care information is subject to strict legal protections, and an RFID system must include good encryption and security to comply—that’s not readily available yet. But there was another issue: Hospital administrators were unwilling to admit the problem existed. They agreed that mixups happen, but only at other hospitals.

So far, Gadh’s biggest success has been tracking parts of cadavers used for medical research and training, to stop them from being lost or sold (see Calif. Researchers Tag Cadavers, Body Parts). Since the body parts belong to people long dead who can’t be personally identified, privacy hasn’t been an issue. And as some of the body parts are worth many thousands of dollars, everyone involved has considered the application a good idea. Gadh says the system is in use at UCLA, and could be made available to other teaching hospitals.

Three interesting conclusions can be drawn from UCLA’s RFID experiences thus far. First, getting the technology to function properly is no longer the hardest part of developing an RFID solution. Today, there are plenty of commercially available tags, readers and other components that skillful designers can use to build functional systems. Second, achieving a return on investment, which used to be the other big challenge, is no longer a hurdle. Those modern RFID components don’t just work, they are cost-effective, too. Third, even if a system delivers undeniable value, some may find it offensive, and political concerns can get in the way. That’s particularly true when RFID is used to track human beings—unless, perhaps, those humans are already dead.

One deployment out of three attempts may not seem like success, but it is, especially when the challenges are not the ones you faced a few years ago. As innovators like Gadh know only too well, progress doesn’t mean having no problems—it means having new problems.

Kevin Ashton was cofounder and executive director of the Auto-ID Center.