WEARABLE COMPUTING

Wearable Communities: Augmenting Social Networks with Wearable

Wearable communities—social networks based on -augmented face-to-face encounters—present both social and technical design challenges. Our WearCom design methodology permits rapid prototyping of wearable community to facilitate community building.

earable computers enhance per- issues and develop a socially acceptable vision of sonal computing with continu- wearable computing. This article presents a frame- ously worn, intelligent assistants work for support for social that augment memory, intellect, networks. Our WearCom wearable community creativity, communication, and design methodology facilitates application creation Wphysical senses.1–3 As wearable computing technol- and provides a framework for investigating the ogy pervades our daily lives, we must ask how it social and technical issues involved. influences social behavior. Will it merely make peo- ple “smarter” by providing seamless, context-aware Wearable communities access to information, or will it also enrich their We use the research framework of computer- social interactions? mediated communities4–6 to investigate wearable To advance the wearable com- computers’ social potential. Wearable communities Gerd Kortuem puting paradigm for personal denote the social networks that might emerge when Lancaster University computing, we must carefully con- enough people use wearable computing technology 7 Zary Segall sider this technology’s social mech- throughout their daily lives. Using a particular tech- University of Oregon anisms, potential, and constraints. nology, namely wearable computers, defines such a Some important research ques- community just as Internet use defines an online tions we must consider include community. Like online communities, wearable communities are multiparty conversations organized • What social potentials do wearable computers around affinities and shared interests, bringing have? together people who don’t necessarily know each • What requirements and characteristics will other personally. Unlike online communities, how- realize these potentials? ever, they are based on embodied, real-world human • How can we systematically and effectively build encounters augmented by wearable computers. such systems? The combination of wearable computing devices • What are this technology’s social impacts? and personal area networks presents inter- esting opportunities.8 WPANs such as The preliminary state of wearable computing enable seamless ad hoc communication over short- technology and the low number of systems in active range radio links (up to 10 meters) and let people’s use keeps us from satisfactorily answering all these wearable computers communicate during face-to- questions. To advance the wearable computing par- face encounters. This permits new forms of sponta- adigm, however, we must start investigating these neous social interactions among collocated people.

1536-1268/03/$17.00 © 2003 IEEE Published by the IEEE CS and IEEE Communications Society PERVASIVEcomputing 71 WEARABLE COMPUTING

Figure 1. Augmenting social space with wearable computers.

online communities have separated physi- cal place and social space, our work on Human interaction wearable communities tries to reunite them—rather than replacing face-to-face interactions, we intend to augment them. For example, when two people convers- ing exchange electronic business cards between their PCs, social interaction leads to digital interaction. On the other hand, inter- actions between wearable devices can facil- Wearable computer Digital interaction Wearable computer itate or even augment social interactions. For example, wearable computers can inform their users about another person’s presence nearby, autonomously exchange information, quietly gather information Unfortunately, wearable computing Touch happens in intimate space, conver- about encounters, or suggest a conversation research generally ignores the crucial inter- sation in personal space. Groups of people topic of mutual interest. They might even play between technology and social behav- set up social distances, usually 10 feet or speculate and inform the user about affinity ior. Current systems and applications less; public space extends beyond this. relationships from repeated encounters.10 emphasize intellectual and sensory capa- Information broadcast by the wearable bilities over interpersonal interactions and computer fills a digital social space whose Applications social competence. The danger of this trend size depends on the wireless transceivers’ To illustrate the interactions that might is that wearable systems may end up transmission range and can range from a take place within a wearable community, we inhibiting interaction. We’ll eventually over- few inches to several feet. The human eye describe several implemented applications. come some of the potentially negative social can’t see the digital field, but the wearable implications that stem from their current computers can. When two computers’ dig- Genie. We designed this simple application state (such as bulky head-mounted dis- ital social spaces overlap, they become to improve knowledge exchange within a plays), but others are inherent to the wear- aware of each other. We see the digital field dispersed group of wearable computer able computing paradigm. By attending to as the alter ego of human social space: just users—for example, students attending the the device and the information it presents as the concept of social space describes fac- same college. Genie makes the group’s com- them, wearable users have less time and tors that influence interpersonal activities, bined expertise available to every group attention to offer people nearby. Such the concept of digital social space describes member. Some might be soccer experts, oth- behavior can negatively influence outsiders’ factors that influence interactions between ers might be hobby gardeners with exten- views of wearable computers and their use. wearable devices. Interactions in the social sive horticultural knowledge; but someone realm can initiate interactions in the digi- who needs to answer a particular question Augmenting social space tal realm, and vice versa. often doesn’t know who can help them. The idea of wearable communities is Based on the notion of augmented social Genie members use their WPAN-equipped based on the notion of augmented social space, we define a wearable community as wearable computers to automatically find space, where WPANs generate a sphere-like a social network that people create or main- fellow community members who can pro- digital field that envelops a wearable com- tain using wearable computing devices. A vide just-in-time expertise. puter and its user (see Figure 1). Social space collection of users becomes a community Using Genie, each user defines a set of defines the space in which humans interact. when enough people use their wearable questions that are stored on the wearable Edward Hall suggests four distinct inter- computers to form webs of personal rela- device. Whenever two or more wearable personal communication proximity lev- tionships. This notion of wearable com- users meet, their computers exchange ques- els—intimate, personal, social, and public munities acknowledges and is based on the tions. This might happen during informal space9—each defining a region around the unique value of random encounters and chance encounters occurring throughout a body with invisible boundaries determined face-to-face interactions. We believe com- day, for example, during lunch or at a cof- by the body’s spatial characteristics, cul- munity building depends on fully embodied fee shop. After receiving a question, the tural conditions, and social relationships. “human moments.” While the Internet and device displays the question and asks the

72 PERVASIVEcomputing http://computer.org/pervasive user whether he or she knows the answer communicate with nearby devices and locations and activities involved. When an and is willing to provide it to the other user. replaces a generic with encounter occurs, the agents produce a Genie relays the response from one com- a more specific presence awareness. negotiation. If both users approve, a deal puter to the other in real time. The system is struck. In a negotiation, the agent eval- transmits only an indication of a person’s Other examples. Genie is a simple appli- uates the value of favors and keeps scores: willingness to engage in a conversation, cation that encourages direct, face-to-face having to run across town just to drop not the answer itself. After a successful interaction. We have also built several off someone’s mail compares unfavorably exchange, the users can approach each applications in which digital interactions with buying milk for someone if the gro- other and discuss the topic of interest face occur implicitly—that is, without users’ cery store is just a block away. Agents to face. To help users identify each other, knowledge but with their consent. employ ideas from game theory to ensure the computers also exchange personal The Piraté collaborative music guide12 that negotiation results are mutually ben- information such as user names or photos. Although simple in many respects, The notion of wearable communities this example illustrates many important aspects of wearable communities. First, acknowledges and builds on the unique value of interactions in wearable communities are situated and involve a rich social context: random encounters and face-to-face interactions. users are aware of whom they are inter- acting with and thus can observe impor- borrows ideas from peer-to-peer file-sharing eficial: they cooperate only if doing so tant social cues such as sex, appearance, applications such as Napster and moves enhances the users’ goals. and gestures. The addition of social con- them to the wearable domain. Piraté lets We envision many other wearable text shapes people’s willingness to engage users exchange MP3 playlists and music communities: with strangers and the particular manner recommendations during brief random in which they interact. encounters. Serendipitously exchanging • Computing communities: Members share Second, wearable communities may or and accumulating information throughout computational resources such as network may not be built on top of existing social the day helps users discover new music bandwidth and computing cycles. networks. In particular, wearable commu- based on the playing habits of those they • Helper communities: Members pledge nities can facilitate face-to-face interactions meet most often, provides awareness of the to assist each other when in distress. between strangers. The only prerequisite community’s favorite music titles, and lets • Bargain-hunter communities: Members for a wearable community is a group of users discover with whom they share a collectively search out sale items in local wearable users willing to cooperate through common taste in music. In contrast to stores. automated interactions of their respective Genie, members of the Piraté community • Marketplace communities: People ex- wearable computers, regardless of pre- do not engage in direct face-to-face inter- change goods without money. existing social ties. actions: exchanges occur implicitly when- • Job market communities: Free applica- Third, you cannot implement wearable ever two or more people come physically tions offer services to passersby. community applications with traditional close. Users aren’t aware of interactions • Knowledge communities: Members col- mobile devices such as and PDAs. between their wearable computers. lectively accumulate information to cre- They require devices that are The WALID (Wearable Augmented ate shared understanding. Task-List Interchange Device) community • Political communities: Members create • Constant: always on and running application digitizes the timeworn tradi- new forms of instant democracy and • Presence-aware: aware of the presence tion of borrowing butter from your neigh- local activism. of nearby devices and people bor.7 You help others because you know • Communicative: able to communicate that one day they will return the favor. From online to wearable with other collocated devices With WALID, two people use their wear- communities • Proactive: able to perform tasks auton- able devices during random encounters to Wearable communities have precursors omously and proactively without requir- negotiate and exchange real-world tasks, in other types of computer-mediated com- ing explicit user intervention (although such as dropping off someone’s dry clean- munities. One of the most powerful fea- interactivity might also be supported) ing, buying stamps at the post office, or tures of today’s Internet is its ability to returning a book. WALID uses per- enable total strangers to interact on a per- This definition is consistent with previous sonal-agent software to find nearby com- sonal and sometimes even intimate level. definitions of wearable computers1–3,11 but munity members and negotiate the ex- People who have never met (and are adds important functional requirements. change of favors. The agents maintain a unlikely to ever do so) use the Internet to Specifically, it emphasizes the ability to user’s task list, becoming fully aware of the discuss personal matters related to health,

JANUARY–MARCH 2003 PERVASIVEcomputing 73 WEARABLE COMPUTING

Supporting Face-to-Face Communication

he notion of wearable communities defines a research REFERENCES framework for investigating computer support for face-to- T 1. CNN, Japan’s Lonely Hearts Find Each Other with “Lovegety,” www6.cnn. face social interactions. Such interactions are often spontaneous com/WORLD/asiapcf/9806/07/fringe/japan.lovegety. (driven by chance encounters of mobile people) and situated 2. L.E. Holmquist, F. Jennica, and J. Wigström, “Supporting Group Collab- (embedded in the real-world context in which people interact). oration with Interpersonal Awareness Devices,” Personal Technologies, Several recently developed systems support such interactions. vol. 3, no. 1, 1999, pp. 13–21. 1 These include match-making technologies (LoveGety ), awareness 3. S. Björk et al., “Pirates! Using the Physical World as a Game Board,” devices to give roaming groups a sense of connectedness (Hum- Proc. 8th IFIP TC.13 Conf. Human-Computer Interaction (Interact 2001), mingBird2), ad hoc games that incorporate real-world group mo- Laxenburg, Austria, 2001; http://viktoria.informatik.gu.se/groups/play/ publications/2001/pirates.interact.pdf. bility into digital entertainment (Pirates,3 Pervasive Clue4), systems for education and exploratory learning (Geney5), collaborative 4. J. Schneider and G. Kortuem, “How to Host a : Support- ing Face-to-Face Interactions in Live-Action Roleplaying,” Proc. Ubicomp 6 augmented-reality systems, and messaging devices that adopt 2001 Workshop Designing Games, 2001; http:// “word-of-mouth” metaphors for proximity-based information www.cs.uoregon.edu/research/wearables/Papers/how2host.pdf. 7 8 9 passing (ThinkingTags, MemeTags, iBalls ). 5. A. Danesh et al., “Geney: Designing a Collaborative Activity for the These systems represent important precursors of wearable com- Palm Handheld Computer,” Proc. Conf. Human Factors in Computing munity systems, yet key questions remain unexplored. First, ex- Systems (CHI 2001), ACM Press, New York, 2001. cept for Richard Borovoy,8 few have investigated this technology’s 6. M. Billinghurst et al., “Experiments with Face-to-Face Collaborative AR social implications over time. Most systems are one-off prototypes Interfaces,” J., vol. 4, no. 2, 2002. that have seen little or no real deployment, so we don’t really 7. R. Borovoy et al., “Things that Blink: Computationally Augmented Name know how augmenting face-to-face interactions affects personal Tags,” IBM Systems J., vol. 35, nos. 3 and 4, 1996; www.research.. com/journal/sj/mit/sectionc/borovoy.html. relationships and communal behavior. Second, we have no sys- tematic engineering process for such systems. Developers have no 8. R. Borovoy et al., “Meme Tags and Community Mirrors: Moving from Conferences to Collaboration,” Proc. ACM Conf. Computer Supported infrastructures or tools to leverage but must build each system Cooperative Work (CSCW’98), ACM Press, New York, 1998, p. 159. from the ground up, in a manner dictated by the underlying net- 9. R. Borovoy et al., “Folk Computing: Revisiting Oral Tradition as a Scaf- work technology and device platform. Many systems thus have a fold for Co-Present Communities, Proc. Conf. Human Factors in Com- limited feature set and cannot evolve to accommodate design puting Systems, 2001; http://web.media.mit.edu/~borovoy/papers/ changes. folk-computing.pdf.

raising kids, romance, and many other top- users. For example, online community sites tionships using mobile phones. Moreover, ics of shared interest. In effect, the Internet such as eBay give users access to commu- the use of traditional mobile devices can provides the connectivity for the “global nity tools from their mobile devices, letting potentially decrease real-world random information community” that J.C.R. Lick- members stay in touch with their commu- encounters and face-to-face interactions: lider envisioned in 1968 as being “not of nity at any time from any place. Also, by attending to people located elsewhere, common location, but of common inter- mobile communities are proliferating in mobile users possibly reduce the amount est.”13 Howard Rheingold, who coined the which interactions between members hap- of time and attention they pay to strangers term “virtual community,”14 described the pen exclusively with mobile devices (mostly nearby. These factors make current mobile Internet as a place that enables a new type through SMS messages and mobile phones). technologies well suited for maintaining of social network, mediated by computer Examples of mobile communities include existing community relationships but less terminals and networks. He defined virtual Upoc, mopilot, and cosmiccupid. apt at creating new ones. communities as “social aggregations that Despite their widespread and successful emerge from the Internet when enough adoption, however, mobile phones and How wearable communities differ people carry on public discussions long Internet-enabled PDAs have serious flaws Like online and mobile communities, enough and with sufficient human feeling as enablers of new communities. They wearable communities consist of webs of to form webs of personal relationships.” don’t support many-to-many communica- relationships that grow from computer- Technologies such as mobile phones and tion very well and don’t facilitate random augmented social interactions. (See the Internet-enabled PDAs have begun to encounters (either virtual or real). Conse- sidebar “Supporting Face-to-Face Com- extend online communities’ reach to mobile quently, it is difficult to create new rela- munication” for current wearable com-

74 PERVASIVEcomputing http://computer.org/pervasive Figure 2. A wearable community system.

munity examples.) They differ from tradi- W tional online communities, however, in three important ways: W W • Social context. Wearable communities are based on direct encounters rather W than indirect, remote interactions. W WPAN exchanges must occur over short distances—when people come face to face or are at least within close physical W W proximity. Communication partners thus can usually see whom they’re inter- W acting with and observe such important social cues as gender, clothing, and ges- W tures. They might even be able to talk to each other. The addition of social con- text shapes people’s willingness to engage with strangers and how they interact. • Usage context. People use wearable individual actions primarily serve their sonal information. devices in different contexts than they own interests. How people deal with con- • Trust. This aspect of human interaction use stationary computers at home or flicting interests determines whether a com- differs when people interact in person work. User attention becomes scarce munity is successful. We’ve identified sev- versus across the Internet. Being able to because, instead of sitting in front of a eral factors that influence people’s behavior look at someone, observe gestures, and computer where they can pay full atten- in wearable communities: mimic them tells you a lot about another tion to the computer and its operation, person. People place much importance wearable computer users also engage in • Identity. Our sense of whom we’re inter- in a handshake when completing a trans- real-world tasks such as driving, oper- acting with shapes all interactions, even action because it signals commitment ating a machine, or simply conversing those with strangers. Our clothes, voices, and mutual trust. No equivalent of a with other people. Time also becomes a bodies, and gestures send messages about handshake exists in online communities. critical resource—when surfing the status, power, and group membership. Internet at home or at work, people have Wearable technology lets people define Wearable community systems more patience with delays because they and broadcast digital identities that Unlike online communities, which have can shift their attention to another task. might not be their true identities. In con- a mature and proven technological founda- This no longer holds true when ex- trast to online communities, these digi- tion, wearable communities require com- changes occur spontaneously between tal identities augment rather than replace plex, less mature technologies such as wear- people who happen to meet during their a person’s true identity, as both are visi- able devices, wireless ad hoc networks, daily activities. ble at the same time. Because observers spontaneous networking mechanisms, and • Technical context. WPANs offer limited can match the digital identity to a per- context technologies. To bring wearable bandwidth and reliability compared to son’s appearance and behavior, people in communities together, we must build reli- wired networks. Piraté showed us that wearable communities can’t easily switch able wearable community systems—collec- we should therefore design systems to identities. tions of wearable devices, distributed soft- exchange low-fidelity data or meta- • Privacy. Unlike security, which deals ware infrastructure, and application information such as song recommenda- with keeping information away from software that let people interact with fellow tions rather than large MP3 files. unauthorized users, privacy gives peo- community members (see Figure 2). System ple the right to control collection and use refers to software and hardware that set the When does a collection of individuals of their personal information, including context for interaction, and wearable indi- become a wearable community? when, how, and to what extent infor- cates the devices’ wearable nature and A community isn’t a perfectly organized mation about them is communicated to implies that they’re always on, commu- group where everyone who participates others. Most wearable communities nicative, proactive, and presence-aware. gains equally. Among other things, people’s require members to disclose certain per- Within a wearable community system,

JANUARY–MARCH 2003 PERVASIVEcomputing 75 WEARABLE COMPUTING

a WPAN that enables seamless connectiv- Successful community design often implementation. WearCom provides ity among collocated devices establishes hinges on what Etienne Wenger calls min- communication. Thus, a wearable com- imalist design,15 which starts with a pro- •A design language that allows the spec- munity system is a highly dynamic, loosely visional design and facilitates community ification of important design decisions connected, potentially large-scale distrib- growth over time. Designers must identify •A design process that outlines an itera- uted system made up of wearable hosts. To the attributes that make communities suc- tive sequence of individual design activ- support the formation of wearable com- cessful and design technology that embod- ities, each of which generates a specific munities anywhere, anytime, such a sys- ies these attributes. When building wear- design artifact (specifications and soft- tem must be independent of external com- able communities, we face the following ware components) munication and computing infrastructures, fundamental problem: • The Proem software platform,18 which relying solely on device capabilities. Sys- supports the implementation and exe- Wearable communities require tem- and application-level software per- hardware and software to support cution of proactive, presence-aware forms essential functions, such as discov- them, but without real-world experi- wearable community applications ering nearby devices and users, initiating ences we do not know what success factors to look for and how to design interactions across wireless links, manag- systems to embody them. WearCom specifies three design elements. ing user identities, and handling user input A user profile—a typed data item that and output. Because of the environment’s Solving the engineering problem and mak- defines a user’s identity within a commu- complexity, developing effective systems ing it easier to develop wearable commu- nity—contains information the user dis- presents a fundamental challenge to build- nity systems will help us create and inves- closes willingly to other community mem- ing wearable communities. tigate these communities. bers along with that deemed useful or To solve this problem, we developed the necessary for community purposes. A com- An exploratory design WearCom design methodology.16 WearCom munity protocol defines how wearable com- approach supports an exploratory design approach puters interact when people meet. Protocols In wearable communities, social con- based on rapid prototyping of wearable deal with communication content, not cerns become difficult to separate from community systems. Rapid prototyping has mechanisms, and in essence formalize ritu- technical practices. To design a wearable proved useful for creating several genera- als of engagement. A community agent— community, we must simultaneously con- tions of the Vu-Man wearable computer.17 an executable software component that sider technology and social interactions Our work extends this by taking a system- implements the community protocol—runs because wearable communities, like all atic approach to designing an entire on the wearable device and functions as communities, result from people’s self- sociotechnical wearable computing system. intermediary between user and community organizing behavior. Relationships emerge The WearCom methodology integrates (see Figure 3). from interactions over time; we cannot social and technical concerns and guides construct them following a simple recipe. designers from scenario development to Case studies We performed several case studies to test our hypothesis that an exploratory design approach facilitates wearable community building. Between spring 2001 and spring 2002 we taught three software engineering courses at the University of Oregon in which students used the WearCom method- ology to design and implement wearable community applications. In total, 35 stu- dents divided into 10 project teams worked on four projects:

Human–agent Human–agent interaction interaction • mClique, a community system designed to improve awareness among mobile users by determining groups (cliques) of Agent–agent interaction mutually connected friends Wearable computer Wearable computer with community with community agent agent Figure 3. Community agents and community protocol in WearCom.

76 PERVASIVEcomputing http://computer.org/pervasive • mBazaar, a wearable community system mass necessary for community building. This together people who have taken on imagi- that brings together buyers and sellers has limited us to highly experimental sys- nary identities based on historical or myth- of goods tems confined to small groups of people at ical characters, and the social interactions • Piraté, a collaborative music guide that our lab and has prevented us from system- contribute much to players’ enjoyment. creates community awareness (discussed atically analyzing the communities’ social Similarly, wearable communities let mem- earlier) dynamics. The projects have, however, given bers invent digital identities to augment • Infomediation, a peer-to-peer informa- us insight into what does and doesn’t work their true identities. Wearable communities tion publishing and access system that when building such communities. should promote such creativity and artistic lets collocated users collaboratively com- expressiveness as much as possible. pare prices at online bookstores Design principles for wearable Principle 5: Use knowledge about past communities encounters to enrich present interactions. WearCom proved successful in guiding Sociologists have extensively researched An important aspect of social encounters is the student teams’ design activities. During the successes and failures of online19,20 and our ability to recognize other people and to the projects, we tracked data relating to face-to-face communities.21 Although we remember if, when, where, and under what learning time, development time, complex- haven’t empirically studied wearable com- circumstances we met someone before. ity of developed applications, and support munity dynamics, we have identified six Wearable computers can track both the requirements. Within 10 weeks, students preliminary design principles that con- user’s current context (defined by other peo- learned to design wearable communities, tribute to successful wearable communi- ple’s presence) and his or her interaction his- implement applications, and experiment ties. Developers can apply these guidelines tory. Applying this knowledge in wearable with the initial prototypes. It typically took to evaluate existing designs, guide the communities can enrich interactions in sig- students less than a week to learn the design process, and educate designers nificant ways. methodology and write their first applica- about the characteristics of successful Principle 6: Include social feedback tion using Proem. After a design phase of wearable community systems. loops. Any community has disruptive or about three weeks, students took five Principle 1: Make users aware of the hid- unruly members—in newsgroups, some weeks using an iterative process to develop den benefits of random encounters. This people consistently and inappropriately application prototypes. Teams deployed principle guided design of the WALID com- flame others; on eBay, some members cheat the applications on a testbed consisting of munity application. When two people meet, by not paying for items they purchase. Suc- , handheld, and wearable devices. they usually aren’t aware of each other’s cessful communities can police their mem- The Proem platform not only simplified activities and, unless they talk about their bers’ behavior not by giving some mem- implementation but also made it possible respective errands, will never find out that bers extraordinary powers but through to run applications on diverse hardware they could benefit from a trade. WALID’s social feedback. For example, eBay uses platforms. community agents look for possible trading reputations (aggregated signed opinions For each project, we tracked support partners and alert their respective users if a from eBay users on a buyer’s or seller’s instances (how many times teachers aided mutually beneficial trade opportunity arises. transaction history) to let members rate students) and, on average, counted fewer Once users become aware of this, they can and judge other members’ trustworthiness. than 3 instances per 10-week course. Con- act on or ignore the agents’ advice. The community thus virtually excludes dis- tributing to this surprisingly low number Principle 2: Reward users for social inter- rupting members. (Jay Schneider22 presents was that we had encouraged students to actions. The frequency of social encounters a good discussion of trust and reputation share their expertise across teams. In sum, per person should be proportional to the in wearable communities.) we found that the WearCom methodology gain a user receives from using the system. effectively supported rapid prototyping of Principle 3: Help people to recognize wearable community systems. each other. Without identity and mutual Despite this success, the overarching goal recognition, a group of people will always of building successful wearable communi- remain strangers. Promoting altruistic illiam Gibson described ties remains elusive. A true community behavior and cooperation necessitates a cyberspace as a “consensual emerges over time and with regular tech- social feedback mechanism that lets users hallucination experienced nology use, and practical considerations lim- evaluate and track others’ behavior. Such a daily by billions of legiti- ited our experiences with the applications feedback mechanism requires identity and mateW operators, in every nation.”23 Wear- developed. A major obstacle is the lack of mutual recognition. able communities present an alternative adequate hardware—wearable computers Principle 4: Support expressiveness. Live- model of human communication that relies that are constant, proactive, communicative, action role-playing games have gained not on shared imagination but on real-world and presence-aware aren’t yet available in widespread international popularity. Such encounters and first-hand experiences. large enough quantities to reach the critical games become social events, bringing Mobile communication and computation

JANUARY–MARCH 2003 PERVASIVEcomputing 77 WEARABLE COMPUTING

technologies have sufficiently permeated our 7. G. Kortuem et al., “When Cyborgs Meet: the AUTHORS society to shape social behavior, norms, and Building Communities of Cooperating Wearable Agents,” Proc. 3rd Int’l Symp. conventions. As mobile computers evolve Wearable Computers, IEEE CS Press, Los Gerd Kortuem is a lecturer in into wearable computers—constant, aware, Alamitos, Calif., 1999, pp. 124–132. at Lancaster communicative, and proactive—we can University, UK. His research 8. IEEE 802.15 Working Group for Wireless expect to see even more dramatic changes. interests include collaborative Personal Area Networks, http://grouper. technologies, ubiquitous We believe that properly designed wearable ieee.org/groups/802/15. computing, wear- able com- computing technology can indeed beget puting, and soft-ware engi- 9. E.T. Hall, The Silent Language, Anchor, community by facilitating social networks neering. He received an MS in Garden City, N.Y., 1973. computer science from the University of Stuttgart based on augmented interactions and face- and a PhD in computer science from the Univer- to-face encounters. 10. M. Terry et al., “Social Net: Using Patterns sity of Oregon. He is a member of the IEEE Com- of Physical Proximity over Time to Infer Wearable communities face sociological puter Society and ACM. Contact him at the Com- Shared Interests,” Proc. 2002 Conf. Human puting Dept., Lancaster Univ., Lancaster, LA1 4YR, as well as technical problems, which are dif- Factors in Computing Systems (CHI 2002), UK; [email protected]. ficult to separate—they interact and coevolve ACM Press, New York, 2002, pp. 816–817. intimately. We plan further empirical re- 11. G. Kortuem, Z. Segall, and M. Bauer, “Con- Zary Segall is the 2002– search on wearable communities and the text-Aware, Adaptive Wearable Computers 2003 Fulbright Distin- social networks this technology creates. as Remote Interfaces to ‘Intelligent’ Envi- guished IT Chair at the ronments,” Proc. 2nd Int’l Symp. Wearable Royal Institute of Technol- Computers (ISWC ’98), IEEE CS Press, Los ogy, Sweden, and profes- Alamitos, Calif., 1998, pp. 58–65. sor of computer science at the University of Oregon. 12. G. Kortuem et al., “When Peer-to-Peer His research has focused Comes Face-to-Face: Collaborative Peer-to- on theoretical and practical methods and sys- ACKNOWLEDGMENTS Peer Computing in Mobile Ad-hoc Net- tems for parallel processing, highly dependable works,” Proc. First Int’l Conf. Peer-to-Peer systems, fault tolerance, real time, binary-to- Computing (P2P ’01), IEEE CS Press, Los This work was performed at and supported by the binary translation, networking, and wearable Alamitos, Calif., 2001, pp. 75–91. University of Oregon’s Computer and Information information systems. He is a fellow of the IEEE Science Department. Computer Society. Contact him at the Com- 13. J.C.R. Licklider, “The Computer as a Com- puter Science Dept., Univ. of Oregon, Eugene, munication Device,” Science and Technol- OR 97405; [email protected]. ogy, Apr. 1968; http://gatekeeper.dec. com/pub/DEC/SRC/publications/taylor/ licklider-taylor.pdf. REFERENCES 14. H. Rheingold, The Virtual Community: 1. T. Starner, Wearable Computing and Homesteading on the Electronic Frontier, Contextual Awareness, doctoral disserta- Perseus Books, New York, 1998, www. tion, Mass. Inst. Technology, Cambridge, rheingold.com/vc/book. 20. P. Kollock, “Design Principles for Online Mass., 1999. Communities,” PC Update, vol. 15, no. 5, 15. E. Wenger, Communities of Practice: Learn- June 1998, pp 58–60. 2. S. Mann, “Wearable Computing: Toward ing, Meaning, and Identity, Cambridge Humanistic Intelligence,” IEEE Intelligent Univ. Press, New York, 1998. 21. E. Ostrom, Governing the Commons: The Systems, vol. 16, no. 3, May/June 2001, Evolution of Institutions for Collective pp. 10–15. 16. G. Kortuem, A Methodology and Software Action, Cambridge Univ. Press, New York, Platform for Building Wearable Commu- 1990. 3. B. Rhodes, “The Wearable Remembrance nities, doctoral dissertation, Computer and Agent: A System for Augmented Information Science Dept., Univ. Oregon, 22. J. Schneider et al., “Disseminating Trust Memory,” Personal Technologies, vol. 1, Eugene, Ore., 2002. Information in Wearable Communities,” no. 4, Dec. 1997, pp. 218–224; www. Personal and Ubiquitous Computing, vol. bradleyrhodes.com/Papers/wear-ra- 17. A. Smailagic et al., “Very Rapid Prototyp- 4, no. 4, Aug. 2000, pp. 245–248. personaltech/index.html. ing of Wearable Computers: A Case Study of VuMan 3 Custom versus Off-the-Shelf 23. W. Gibson, Neuromancer, Ace Books, New 4. N.K. Baym, “The Emergence of Design Methodologies,” J. Design Automa- York, 1984. Community in Computer-Mediated tion for Embedded Systems, vol. 3, nos. 2/3, Communication,” CyberSociety, S. Jones, 1998, pp. 217–230. ed., Sage Publications, Newbury Park, Calif., 1995, pp. 35–68. 18. G. Kortuem, “Proem: A Plat- form for Mobile Peer-to-Peer Computing,” 5. P. Kollock and M. Smith, eds., Communi- to be published in ACM SIGMOBILE ties in Cyberspace, Univ. Calif. Press, Ewing, and Communications N.J., 1998. Rev., 2003.

6. J. Preece, Online Communities: Designing 19. M. Godwin, “Nine Principles for Making For more information on this or any other comput- Usability, Supporting Sociability, John Virtual Communities Work,” Wired, vol. ing topic, please visit our at http:// Wiley & Sons, New York, 2000. 2.06, June 1994, pp. 72–73. computer.org/publications/dlib.

78 PERVASIVEcomputing http://computer.org/pervasive