Future of social networking explored in UW's computer science building

Research members have been testing the system on themselves. Over the past year, all 10 members of the group have worn the tags on their necks and placed them on certain belongings. Balazinska set the system so that she can't see her students, but she allows them to access her data. The students occasionally used the alerts to catch their adviser on her way out of the building. Many members of the team reported using the database to find out where they had left their belongings.

The pilot study will incorporate two new features, developed by students, that aim to exploit the system's potential benefits. One invention is a tool that records a person's movements in Google Calendar. Study participants can set the system to instantaneously publish activities on their Web calendar, such as arrival at work, meetings or lunch breaks.

"It's a perfect memory system that records all your personal interactions throughout the day," Welbourne said. "You can go back a day later, a month later, and see, ‘What did I do that day"’ or, 'Who have I spent my time with lately"'"

Another tool is a friend finder, named RFIDder (pronounced "fritter"). This sends instant alerts to participants' e-mail addresses or cell phones telling them when friends are in certain places. With RFIDder, each user can specify who is allowed to see his or her data. Users can change the settings at any time, and can easily turn it off whenever they don't want to be found. The system will link to Twitter, an online blog that lets people post their whereabouts online.

"We want to observe how a group of people uses these tools, whether they find them useful, how they adapt them," Balazinska said.

Researchers are also devising ways to deal with the many technical challenges involved in sorting RFID data. As data floods in, the researchers would like to make sense of it. They also want to develop a main database where people can find the information they need, but can't abuse it by looking at too much of other people's personal information. Proposals include systems that would impose a cost for looking up certain types of information, or that would let people see who is accessing their data.

A major research focus is extracting information from imperfect data. Metal can block the RFID signal and using the data to figure out people's actual position is tricky. Current systems combine artificial intelligence and database techniques to produce usable information, Balazinska said.

"This is a major project that has many facets," Balazinska said. "We worry that these technologies are being implemented too quickly, and with this system we want to explore it in a controlled environment, to inform the public and policymakers about issues we might face."

The project is funded by the National Science Foundation, Microsoft Research and the UW's College of Engineering.

For more information, contact Balazinska at (206) 616-1069 or ; Borriello at (206) 685-9432 or ; and Welbourne at (206) 276-7969 or .

More information on the RFID Ecosystem is at http://rfid.cs.washington.edu/.