Issue No. 01 - Jan.-Mar. (2013 vol. 12)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/MPRV.2013.18
Nigel Davies , Lancaster University
Pervasive and mobile computing differs fundamentally from topics such as operating systems and networks. While the latter can be considered horizontal layers with clearly defined boundaries (at least in theory), the issues of ubiquity and mobility touch on all aspects of a system and its use. For example, coping with the resource constraints inherent in a mobile device impact all aspects of a system, from the hardware design and device's form factor through to its user interface and the amount of attention that the average user can spend interacting with applications. It's the all-encompassing nature of the challenges presented by mobility and ubiquity that make the field so interesting.
In a similar vein, there are issues that affect all aspects of a mobile and ubiquitous system. For example, tackling privacy concerns might require specialized hardware, new network protocols and systems software, and applications and user interfaces that present users with appropriate tradeoffs and offer a user-education program. Over the next few issues of IEEE Pervasive Computing, I hope to discuss several of these cross-cutting issues—starting with an exploration of time in ubiquitous systems.
Time is often the forgotten form of context information in mobile and ubiquitous systems. As researchers began to explore how to build systems that adapted to context, location was the "new kid on the block," and designers rushed to build location-based systems. As our understanding of location matured, new forms of context—such as device orientation, user preferences, and physical activity—became important. However, in many ways, time is one of the most important forms of context. From sorting search results for shops based on opening times, to dynamically creating city tours optimized for different times of day, to creating user interfaces customized for users with limited attention, using temporal information to tailor our experience is crucial. Users will likely reject systems that sound alarms at inappropriate times of the day or recommend dinner options in the morning.
Time also has an important role to play in how we capture and interpret context. For example, recognizing activities such as chopping vegetables within the broader context of preparing a meal is significantly easier than recognizing the activity in isolation. However, it's not just the order that's important—knowing the time of day during which an activity is being conducted can significantly improve recognition rates. Indeed, areas such as activity recognition and formulation of adaptation rules require researchers to develop sophisticated techniques for describing time and the temporal relationship between events.
At a systems level, time drives many of the protocols and system processes that underpin modern ubiquitous systems. The effort to reduce power consumption—from refinements to duty-cycle strategies—is a prime example of how we can deliver energy efficiencies by sacrificing outright temporal performance.
Although system performance is typically measured in nano- or milliseconds, ubiquitous computing also requires us to think in much longer time scales. Many modern computer systems are designed to have a short lifespan, and consumers frequently change their mobile devices. However, to realize the vision of ubiquitous computing, we must embed computers into everyday objects and the built environment. This fundamentally challenges the way we design our systems, because the systems will need lifespans measured in decades or centuries, not weeks or months.
Perhaps the most interesting use of time comes when we begin to consider how ubiquitous systems can change human perception of time and its impact on how we structure our lives. Using technologies such as Microsoft's SenseCam, 1 mobile systems can capture much of the detail of our everyday lives. The data captured can then be replayed at a later time to provide new insights into our past. Providing highly detailed records of the past, available for immediate recall, will likely significantly change how we perceive the passing of time and plan for the future.
Of course, repurposing devices that we associate with time for other purposes also has a long tradition in ubicomp. Back in 2000, IBM developed the Linux Watch in an effort to make significant computing capabilities available to users on their wrists. 2 Now, hardware with such capabilities is relatively common, although user uptake has been slow. The Whereabouts Clock also provides an interesting glimpse into how we might combine location and time to provide totally new forms of information, presented in the context of a familiar setting (in this case, a clock). 3
In this issue, we focus on the role of mobile and pervasive computing in transit. I thank the guest editors, Albrecht Schmidt, Rajesh Krishna Balan, and Brian Ferris, for putting together this exciting issue. Owing to environmental and economic concerns, finding new ways to make our transport systems more sustainable is immensely important, and the three theme articles in this issue do an excellent job of highlighting related opportunities and challenges.
Time is also an important factor in transport. Current travel plans are dominated by a strict notion of time, and the need to conform to schedules leads to increased pressures for travelers and to inefficiencies when these schedules can't be met. However, the tight coupling between transport and time is a relatively recent event. For example, it wasn't until 1880 that legislation enabled the UK to run in one unified time, Greenwich Mean Time (GMT). Prior to this, towns and cities across England, Scotland, and Wales ran on their own time, derived from the solar day. However, the advent of the railway network and the growth of the telegraph system meant that by the mid-19th century, major cities, such as Liverpool and Manchester, began to run on London time. Before long, timetables and station clocks across the UK began to shift to GMT, as more trade and communication began to rely on the new technologies of the age.
Mobile technologies provide us with new and interesting ways of relating time and transport. For example, my research group at Lancaster University is working with a highly interdisciplinary group of transport specialists, psychologists, geographers, and digital artists. We're exploring how we can leverage mobile technologies to offer users different ways to relate to time and understand the relationships between their transport plans and those of others by highlighting future intersections of the trajectories of people, objects in transit, and modes of transport. 4
We've started prototyping solutions in the tourism, logistics, and school transport domains, attempting to help decouple our notions of time to support more opportunistic and sustainable forms of transportation. To help convey the project's ideas regarding time, we use storyboards for user engagement (see an example in Figure 1). The storyboards offer insight into time-based experiences that are difficult to articulate without a narrative that includes a series of participants and stakeholders.
Feature Articles And Departments
This issue also presents four feature articles on mobile and pervasive systems. In "Building a Low-Cost Digital Garden as a Telecom Lab Exercise," the authors describe an innovative, interdisciplinary, semester-long class in which undergraduate students learn how to build a low-cost, multihop wireless sensor network for detecting soil moisture and environmental temperature.
Moving from the garden to the home, the next article, "A Smarter Smart Home: Case Studies of Ambient Intelligence," aims to optimize home use with computational interventions. The authors target behavior change for sustainability, which should be of interest to readers concerned with either smart environments or sustainability in ubiquitous systems.
IEEE Pervasive Computing spans the full range of research, from networking and systems to users and social impact, and in "Self-Organizing Seamless Multimedia Streaming in Dense Manets," the authors describe a set of protocols for automatically determining dense areas of interest and seamless support of multimedia streaming in pervasive manet environments.
Finally, in "Mobile Biometrics: Combined Face and Voice Verification for a Mobile Platform," the authors consider how to combine real-time face and voice verification to improve the security of access to personal data from mobile phones.
We also continue to strengthen our departmental offerings, which remain one of the most popular parts of the magazine. Elizabeth Belding has taken over the Conferences department, which reports on the recent Ubicomp 2012 conference. Our new department, "Notes from the Community," edited by Jason Hong and Mary Baker, gets off to a flying start, drawing material from the pervasive computing subreddit (which anyone can join at www.reddit.com/r/pervasivecomputing). We've also relaunched our Pervasive Health department, with Anind Dey, Jesus Favela, and Stephen Intille as the editors. In this issue, the department explores the use of wearable cameras to help capture travel behavior to encourage travelers to select more sustainable and active travel modes. In addition, two of our departments, Innovations in Ubicomp Products as well as Smartphones, explore how pervasive displays will form part of our future ubicomp environments.
Finally, I'd like to welcome Sunny Consolvo to the IEEE PervasiveComputing editorial board. Consolvo is a User Experience Researcher at Google. She focuses on persuasive technologies, ubiquitous computing systems, mobile computing, and Web technologies (see the sidebar).
From transport to health, from understanding user behavior to exploring data security, and from new network protocols to new uses for smartphones, I truly hope there's something for everyone in this issue.
The Sixth Sense Transport project ( www.sixthsensetransport.com) is supported by an Energy/Digital Economy UK Research Councils grant.