Pages: pp. 2-3
One of my earliest involvements with IEEE Pervasive Computing was writing an article for the inaugural issue on the challenges of deploying ubiquitous computing systems. 1 The article focused on the technical and sociological challenges of creating systems that are ubiquitous in the more general sense of the word—systems that extend beyond mere laboratory prototypes and become an integrated part of our everyday lives. A central tenant of the article was that despite the technical advances that had occurred in the 10 years between Mark Weiser's article, "The Computer for the 21st Century," and the first issue of IEEE Pervasive Computing in 2002, many of the use cases described in his article still appeared as science fiction precisely because the systems hadn't yet moved out of the laboratory and into the "real" world. 2
The issue of deployments is one I've returned to several times in my own career. The benefits of deploying systems, either to reap actual benefit or simply as part of an evaluation process, are numerous. However, the difficulties associated with such deployments are also numerous, especially when the deployments take place in public. 3 For some time, there has been an impasse—as a community, we know that large-scale deployments are useful but their cost and complexity often mean that we have to make do with lab-based experiments. Recently, however, the landscape has changed significantly with the emergence of deployment models based on application stores, such as the iTunes app store. Using such deployment methods, it's possible to create applications that can be very widely deployed with fairly minimal effort. A recent workshop explored this topic in detail and reported on several research projects that had used app stores for deployment or evaluation purposes ( http://large.mobilelifecentre.org/dotclear/index.php?post/2010/05/11/Workshop-at-Ubicomp-2010).
Using app stores to support experiments comes with obvious difficulties—for example, a lack of control and an inability to capture the types of demographic data normally associated with an experiment—but such stores bring huge benefits as well, such as the increased speed and reduced cost of deployments. This in turn facilitates experiments in areas such as social computing and large-scale opportunistic sensing (a topic we'll cover in a future issue; see www.computer.org/pervasive/cfp4 if you'd like to submit an article for peer review) as it makes it possible to get applications into many users' hands at a very low cost.
App stores and similar models work well for mobile devices, but obviously have limitations when physical installations are required—for example, when investigating smart environments. In such cases, we're starting to see a drive toward collaborative projects and open initiatives that aim to amortize the cost of deployments over multiple experiments. One example is the Ubiquitous Oulu initiative ( www.ubioulu.fi). The work in Oulu, Finland, is aiming to systematically explore the use of ubiquitous computing technologies in an urban environment; researchers have created a rich distributed infrastructure of wireless networks, sensors, and public displays in the city. Through a series of challenges, researchers from outside the project are encouraged to submit proposals for applications that can run on this infrastructure—creating all sorts of possibilities for mutually beneficial research projects.
In my own work, I've been collaborating with my colleagues to unite these two ideas into a model in which app stores can be used as part of a deployment process for ubiquitous computing applications for large-scale public display networks ( http://large.mobilelifecentre.org/dotclear/public/Davies-in-the-large.pdf). This work is being carried out in the context of the PD-NET project ( http://pd-net.org), which is exploring the creation of an open global display network and has already created a testbed of displays with nodes in four European countries (the UK, Germany, Switzerland, and Portugal). The motivation for trying to combine app store models of deployment with innovative and scarce large-scale deployments is much the same as my motivation for writing that initial paper in 2002—to try to provide tools and information that help us move ubiquitous computing from the laboratory to the real world . One way in which IEEE Pervasive Computing can help further this vision and fulfill its role as a catalyst is by making the community aware of the opportunities to share infrastructure. If you have a particular initiative in this area that you would like us to cover, please drop us an email or, alternatively, post a pointer on our Facebook page for the benefit of all.
I thank our guest editors Joe Paradiso, Prabal Dutta, Hans Gellersen, and Eve Schooler for their work creating such a strong issue on a particularly important and timely topic—smart energy systems. In addition to the theme articles, this issue also features several articles that cover a broad range of topics. In "Ubiquitous Advertising: The Killer Application for the 21st Century," John Krumm describes how ubiquitous computing research can help support advertisers in areas such as ad targeting and feedback. This article also serves as a good introduction to topics covered in our forthcoming issue on pervasive retail, due to be published in the following issue. Nicholas K. Taylor and his colleagues describe the Daidalos project in "Pervasive Computing in Daidalos," focusing in particular on systems/middleware support for user access to services. Finally, in "Detecting Fall Risk Factors for Toddlers," Hana Na, Sheng-Feng Qin, and David K. Wright focus on a smart vision system designed to assist in toddler supervision. The system tries to preempt injuries by detecting risk factors and alerting caregivers before a fall takes place.
Last but not least, I'm absolutely delighted to announce that we have two new associate editors in chief (AEICs). Both Maria R. Ebling and Albrecht Schmidt have agreed to serve as AEICs for the magazine. Ebling has had a long association with IEEE Pervasive Computing, serving on the editorial board and most recently editing the New Products department. Schmidt is serving on the editorial board and is well-known in the ubiquitous computing community, having chaired many of the major conferences in the field. I'm looking forward to drawing on the expertise of both of the new AEICs, and I believe these appointments will really help us move the magazine forward.
Maria R. Ebling is a research staff member and senior manager at the IBM T.J. Watson Research Center. Ebling has a PhD in computer science from Carnegie Mellon University. Her interests are in distributed systems supporting mobile and pervasive computing, privacy, and human-computer interaction.
Albrecht Schmidt is a full professor of pervasive computing and user interface engineering at the University of Duisburg-Essen. He has a PhD in computer science from Lancaster University. His current work focuses on novel user interfaces and interactive applications in the context of mobile and ubiquitous computing.