Yesterday was my first day at Nokia Research Center, Palo Alto, and today was my first day at a ubiquitous computing conference since UbiComp 2003 (for which I was General Chair). It feels like a homecoming of sorts -- reconnecting with many people I haven't seen in a long time, and finding many themes being presented that are [still] of great interest to me.
Crista Lopes, Paul Dourish, Adrian Friday and the rest of the conference and program committees have done a great job this year. The paper acceptance rate was only 13% (30 of 232 submissions), and the quality seems pretty high, among the papers I saw presented (which was only half, as the conference is experimenting with two tracks this year). I was glad to see a large number of interesting demos and posters as well.
Bruce Sterling, author of Shaping Things (among other books), led things off, saying that what appealed to him about ubiquitous computing was "the majesty of the ideas and the lyricism of the language". His "pencilpoint" presentation was a call to action for ubiquitous computing in service to sustainability, sketching out a SPIME (= SPace-tIME) Meme Map that emphasized thinking, designing, naming, making, tracking, searching and recycling in the coming Internet of Things, where all physical artifacts will have associated digital representations. After making the case for SPIMES, and the need for design for sustainability, he railed against artificial intelligence, panpsychism, magic, fundamentalism and fanaticism. I think his book is great, but -- perhaps because I'd already read the book -- the presentation was not as inspiring as I'd hoped ... and I didn't appreciate the way he treated questions, and questioners.
Ryan Aipperspach of Berkeley Institute of Design and Intel Research Berkeley presented "A Quantitative Method for Revealing and Comparing Places in the Home", in which he adddressed the question what is a place? He distinguished between positional features and behavioral or historical features of -- or at -- a location, and presented an algorithm for applying gaussian mixture models in a three stage process to cluster positional data of people collected via ultra wide band (UWB) sensors. The location traces were made from 3 homes over periods ranging from 2 weeks to 3 months. Future work may include larger studies, combining the quantitative data with qualitative data, and perhaps articulating more compelling value propositions for using technology to identify places in the home.
Scott Davidoff, of Carnegie Mellon University, talked about "Principles of Smart Home Control", in which he raised the issue of what is control? He distinguished the technically-oriented control of devices from the socially or anthropological control over lives, and observed that most dual-income families really want is to achieve mastery of busyness, rather than [simply] managing time and activities, and proposed using ubiquitous computing technology to support routine tasks to allow family members to participate more fully in activities from which they derive meaning (though Bill noted, in an aside, that any activity can be infused with meaning ... there are "no ordinary moments"). Scott's observations about the incremental and improvisational planning in which many families engage reminded me of research done in artificial intelligence in the domain of planning.
Susan Wyche, from Georgia Tech, presented research on "Historical Analysis: Using the Past to Design the Future", in which she asked how are we going to clean the [smart] home of the future? Exploring yesterday's future helps defamiliarize the present, and better recognize the insights history has to offer. The historical analysis was directed at technology adoption in U.S. homes from 1920 thru 1965, with a focus on household chores. Among the most interesting findings (to me) was that technology often results in reductions in force (e.g., using a clothes dryer or dishwasher), but a concommitent reduction in unplanned interactions and togetherness (chatting with a neighbor while hanging clothes out to dry or with a family member while washing dishes by hand).
Sri Hastuti Kurniawan, from the University of Manchester, presented "An Exploratory Study of How Older Women Use Mobile Phones" offered an interesting twist on the grandmother test. She uncovered some age- and gender-related biases in preferences for mobile phones among older women. Size matters, but the size of the text [font] is more important than the display size, and bulky phones are regarded positively as they are easier to grasp and hold. Fear and security are key motivations behind mobile phone purchases and usage in this group. Happy slapping -- assaults allegedly perpetrated simply for the purpose of capturing a photo of the attack -- was singled out as a source of that fear ... and Bill noted that happy slapping is primarily a media-hyped hoax (another reminder that no one can terrorize us without our consent).
Julie Kientz, of Georgia Tech, talked about "Farther Than You May Think: An Empirical Investigation of the Proximity of Users to their Mobile Phones" [slides]. She and her colleagues wanted to determine whether assumptions about mobile phones always being within arms length are correct. They conducted a field study in which 15 people (9 female,s 7 males, ages 21-65, with monthly usage patterns ranging from 4500 minutes down to 25 minutes) wore Bluelon Bluetooth tags for 24 hours / day over 3 weeks period. The tags recorded an estimate of how far away the user's mobile was once every minute, and showed that, on average, mobiles are within arm's length (2-3 feet) 50% of the time, in the same room 18% of the time, not available 20% of the time and off 12% of the time. Weekly interviews were used to compliment the quantitative measurements, and it was interesting to see that many of the participants also assumed [incorrectly] that their phones were always within arms length.
Beki Grinter and Ken Anderson volunteered for an impromptu panel in lieu of a presentation of a paper whose author did not attend the conference. They discussed some of the problems inherent in ethnographic studies of the use of an artifact that has gone where few artifacts have gone before (Beki started out asking how many people had overheard mobile phone conversations by people in bathroom stalls). Among the many interesting audience questions and comments was Marc Davis' observation of the tension between quantitative and qualitative methods, and the claim that we are on the bverge of computational social science, where quantitative methods can help address problems in qualitative methods such as scale and memory fallibility. Someone (whose name I did not catch) then said that, in his experience, there is an iterative process whereby quantitative methods can help provide focus for qualitative methods, which can then identify new dimensions in which to apply quantitative methods ... iterate as necessary.
Having been away from UbiComp for 3 years, it is interesting to see how much of the research [presented today [in the sessions I attended]] is focused on mobile phones. Given my new affiliation with Nokia Research Center Palo Alto, I was glad to see that Nokia is a conference sponsor, and that another member of NRC, Cristiano di Flora, organized a workshop on System Support for FUture Mobile Computing Applications (FUMCA 2006) that was part of the workshop program at the conference. I am also heartened to see many references to the use of Nokia phones in the field studies, applications, demonstrations, and concepts being presented here. However, there were no papers (or demos or posters) at the main conference from Nokia ... hopefully we can change that for next year's conference.
Sounds like a good time, Joe...sorta. Congrats on the new gig!
Posted by: adam | September 20, 2006 at 08:18 AM