My research is in the area of ubiquitous computing, or “ubicomp.” I’ve been working on applications for mobile phones, particularly applications that help people to stay socially connected in the urban computing environment. My research focuses mainly on building mobile applications that will improve people’s lives without violating their privacy. As a company, Intel takes a strong stance on the importance of protecting the user’s privacy.

In my research, I collaborate with researchers at Georgia Tech, Carnegie Mellon and the University of Washington. We’ve spent a lot of time exploring what tasks users are trying to accomplish with mobile phones, so we can build applications that make it easier to accomplish the tasks.

A classic example is the use of mobile phones to coordinate physical meetings. For instance, we know from research by cultural anthropologist Mimi Ito that in Japan, teenagers do not make precise social plans. They might say “we’re going to be in this neighborhood at about this time.” Then they use their mobile phones to do a complicated exercise involving many people sending text messages and making phone calls to coordinate where to meet. We’re trying to develop an application that would make it easier and less awkward to do that, while protecting the privacy of each person in the group.

There are a number of techniques for protecting privacy in a ubicomp environment. I think the single best approach is to ensure that users maintain control of their own data. That’s why the techniques we use to compute the location of a device are run on the device itself. They don’t use the network, so users don’t have to trust a third party to do this work.

Another way to protect privacy when people exchange information is to let each user decide what information to disclose. For example, rather than tracking a person’s location remotely, our technology gives each user control over what information to disclose. Under our approach, a group of people could agree on a place to meet without invading the privacy of anyone in the group. I don’t have to tell you where I am in an effort to agree to meet somewhere.

Another task that’s awkward to carry out with a mobile phone is letting a friend know when you’re running late, or when you will arrive at a certain location. To avoid the need to make periodic calls with updates, one application we’re exploring would enable your cell phone to continuously transmit your location to your friend’s phone. Your friend could simply click a button on her phone to get status updates. The application could be designed to shut itself off automatically when you get close to your destination, so that you don’t accidentally leave it on and end up violating your privacy. You could make the application even more privacy-sensitive by simply providing an estimate of how long it will be until you arrive at her location. This doesn’t violate your privacy as much, because it doesn’t reveal your location.

Another simple application prototype we developed would enable users to click a button on their location-enabled phones and tell their family and friends where they are. When the button is clicked, the phone would send a brief text message to family members or friends, such as “John is at home,” or “John is at the office.”

We implemented a number of small field deployments in 2005 to test a variety of ideas for ubiquitous computing applications for mobile phones. We’re now at a point in the research where we are investigating the potential to translate our technology into products, within Intel or externally. We know from our research that there’s a great need for ubiquitous computing applications to support people in their social lives. These applications could greatly expand what people can accomplish with mobile devices, without the need to give up their privacy.

Ian Smith joined Intel Research Seattle in 2003. His research focuses on ubiquitous computing technology, social science, and software engineering, with an emphasis on the privacy issues related to location-enhanced computing. Prior to joining Intel, Smith was a researcher at the Palo Alto Research Center (PARC). He earned his Ph.D. in 1998 from the Georgia Institute of Technology.

Smith serves on numerous program committees and is a reviewer for several leading computer science journals and conferences. He is frequently invited to speak at universities and at industry events on topics such as privacy and new usage models for mobile phones.

Smith is the author of numerous conference papers, journal articles and book chapters. Most recently, his research into mobile technologies was presented at CHI 2005, Pervasive 2005 and Ubicomp 2005 and his work will be featured at the 2006 versions of these venues as well. He holds two U.S. patents and has ten patents pending.