ambient intelligence : mas961

project update

2005-04-13T10:08:00.000-05:00

[a complete turnaround]

scope, well-revised

well, i've decided to go the capricious route and completely change my project design. 'submergence' had its fans, yes, but i've chosen to submit that to the ars electronica next idea application and allow it to simmer within the throes of conceptual art. the heart-rate thing seemed to be too close for comfort and, incidentally, i haven't had much luck procuring affordable heart rate monitors to tinker with.

the next next idea: quilted pixels. i'm fascinated with custom fabricated textiles + clothing, and moving the invisible digital back into the physical realm. i've been playing with this idea of fabric that is not decorated externally (i.e. silkscreening, iron-on, stitching), but rather has its decoration potential installed internally, as an all-in-one package. the material itself could provide creative resource and decorative possibility to the wearer; the control shifts from the manufacturer and choice-limiter to the consumer as creator.

scenario plays as such: one purchases a stash of 'blank' apparel or accessories, such as one would get white paper. an image could be acquired by the user, either by directly downloading an image to the garment (akin to choosing + ordering a specific ringtone for your phone), capturing an image through a camera (akin to recording a real-time sound as a ringtone), or creating your own image through a drawing pixel program (akin to creation via ringtone composer). after specifying the image + position + color (in an ideal scnario), the wearer inputs the data into the garment and the fabric then dyes itself through pixels that bleed the appropriate colors.

a strong point of this design is that the image would be permanently stained onto the clothing. the fabric isn't merely a dynamic display; it's a conscious action to 'develop' such an image onto oneself, a decisive branding. the color permanence gives it weight + identity significance.

hard stuff: by using quilted fabric, each pixel will hold some sort of dye that will release upon trigger. still trying to figure out the best way to do the color transfer. some ideas: (1) putting dye into capsules (stage-blood-esque, or very weak paintballs) and triggering via physical breakage (electronic pinches?) or a motored grind; (2) using a central reservoir which flows/leaks dye into corresponding pixels by controlling the valves; (3) using thermochromic ink, which changes color with heat. i would want an irreversible type of this dye.

soft stuff: right now so far

fabric: got gobs of white cotton quilted fabric. each pillowed pixel is about half-inch square. research: recipes for stage-blood, thoughts of using gelatine capsules, heating elements, thermochromic dye (colorchange.com, but the website makes your eyes bleed), the ruug project (http://www.andrew.cmu.edu/user/adrianat/portfolio.html), how high of a resolution needed to convey image without being totally unwieldy (a 5x6 matrix?).

planning ahead

at the very least, would want my own hard-coded designs to appear. would like to make a simple GUI software that would be a pixel-paint program that would enable users to design + download their created designs.

augmented shoes

2005-04-05T23:04:00.000-05:00

[twistings]

inspired by this last weekend’s less-than-ideal encounter with the latest spring fashions, i’ve been thinking of ways to embellish and revolutionize the often revered but easily overlooked wearable accessory: the shoe.

in relation to how many choices there are in clothing sizes and silhouettes, most shoes are limited in terms of last (the shape of the foot inside the shoe) and volumetric proportion (foot thickness, or ratio of ball to heel width). save for expensive options like specialty shoes or marginal sizes, human feet are often squeezed into ill-fitting options. since there exists such variable sizing standards among companies and also ‘give’ of certain materials and fabrics, shoe shopping is tedious, imperfect, and usually requires in-person try-out (“if the shoe fits, wear it”). one solution to this might be having a foot scanner that models a person’s physical appendage (this already sort of exists for jeans at selfridges from bodymetrics) and then recommends certain sizes and shapes for the wearer; this way someone could even feel comfortable ordering shoes from a catalog or online. another solution might be to insert sensors into the interior of an existing shoe. either the sensors can be used during the shoe-fitting process (if you try on a shoe, sensors will indicate exactly where tightness or looseness occurs, and recommend a better size), or used within existing footwear to collect data over time (like flatfootedness or necessary heel cushioning) in regards to how a person can obtain shoes that are optimally comfortable and durable.

other thoughts on shoes: it might be helpful if shoes could improve the surface on which they stepped. perhaps there could be a sensor on a shoe that would recognize if it were indoors, and slide an indoors sheath or repellent material along the bottom; the outdoors and indoors sole contamination would be minimized. also, perhaps a cleaning sole could be attached or dispensed so that someone could help keep their floors clean by merely walking about their home over time. toss your swiffer! other ideas would be shoes with built-in pedometers and pressure sensors to record and distinguish low-impact and high-impact activities, an RFID bump-toe interface (input through tapping or kicking an object), and customization via multiple styles of conductive lacings through the eyelets.

augmented reality

2005-03-19T17:40:00.000-05:00

[papers]

Augmented Reality: A New Way of Seeing, Steven K. Feiner, 2002
http://www.sciam.com/article.cfm?articleID=0006378C-CDE1-1CC6-B4A8809EC588EEDF

[comments + questions]

i can clearly see the value in augmented reality perfectly working... the classic examples of the mechanic trying to identify parts + processes, the doctor operating on surgery, someone wishing for a directly mapped navigation. having the information we need when + where we need it would be astonishingly useful, but i think the dream falls short when dealing with the clunky head-worn implementation and troublesome episodes of too-much or too-few or too-irrelevant items of information. until computers truly know what we're thinking or intending, you fear the risk of introducing distracting clutter. the paper doesn't directly address the user interface, namely turning on or off parts of the augmented view, so i'm not exactly sure how the user specifies what to display and when. of course, there's the choice of removing the head display completely to turn off the thing...
augmented reality might be good for the individual user in times of displaying helpful information, a selfish sort of functionality, but what happens when you start considering it in the context of social culture? [perhaps engineers don't put this high on the priority list, but...] one of the first scenarios i envisioned was a hopeless lothario-wannabe, who had pickup-lines and love-tips fly into sight while out on a dinner date. the reminders or suggestions ("compliment her outfit, stupid!") would prettily augment her forehead or other well-displayed body parts. although this is a silly, lame example, you might wonder what other people might be augmenting *on* or *what* when they're talking to you. are they projecting their secret pet peeves about you? are they projecting some twisted fantasy, or a banal grocery list? mapping personal information [practically] directly onto another individual seems kind of creepy and inappropriate. it's the same discomfort one feels when approached by someone behind a large camera, or wielding a pair of binoculars. the intermediate object, however benign, dramatically interferes with face-to-face interaction. if the tool becomes invisible or imperceptible, distrust may brew [which we can see today as many establishments now ban camera phones].
maybe a more subdued version of AR would be a tool to virtually annotate the physical environment. it would be more of a personal sketchbook or notepad instead of general reference information. instead of seeing a restaurant with all its reviews virtually augmenting its facade (a weird point in which most restaurants do poster their windows with articles and reviews), one could post their own personal experience or dining review onto the restaurant, which would pop up whenever you returned to the establishment. the annotation tool would be more 'the world is your sketchpad' than 'the world is your television'.

tangible + ambient

2005-03-12T13:42:00.000-05:00

[papers]

Tangible Bits, Ishii + Ullmer, CHI97 [presentation] [videos]
check out website Tangible Media Group

“Slow Technology”, Hallnas & Redstrom

Personal + Ubiquitous Computing Journal Vol 8 Nr 5 Special Issue on Tangible Computing

Ambient Interfaces: Design Challenges and Recommendations, Fraunhofer Institute

Cohen & McGee, Tangible Multi-modal Interfaces

Mobile Feelings, Sommerer + Laurent Mignonneau

[comments + questions]

the article i found most appealing and pertinent was the 'slow technology' paper by researchers at the interactive institute in sweden. i sympathize with the motivation behind their work; namely, technology that may not be wholly instanteously gratifying, where you 'get it' not upon immediate use but with extensive and patient degrees of time. the flashy demo + gadgety stuff superficially impresses people, rather than changes their perspectives in interesting and unpredictable ways. can we look to nature for the beauty in slow, reflective processes? seasons change gradually [a harsh winter particularly seems to last forever], yet with each passing year we cry, 'where did all the time go?' there exists a pervasive trend in our culture that looks speedily forward but never stops to treasure the present. technology, with all its power, seems to be primarily utilized to compress time unnaturally, spinning into a perpetual catch 22; i hope that we can balance it out with technology that expands time poetically and artistically. as we appreciate more the everyday tasks and tools, the easily-forgotten yet inarguably-deep associations between ourselves and other people and places garner authenticity and significance.
prof. ishii's philosophy manifests itself in seamless computing, where the world becomes the environmental interface, in line with mark weiser's vision of true ubiquitous computing. however, many of the research projects within tmg deal with highly specific task-based tools, harnessed to a flat, finite tabletop augmented with tagged phicons, a projected image, and various cameras and sensors. although touching actual objects has its tangible satisfaction, the table interfaces still seem very much two-dimensional, digitally expressed, and artificially implemented. however, i applaud the non-table projects, such as topobo, comtouch, and the i/o brush. tangible media seems to be most favorably effective when the technology becomes incorporated within, not augmented upon, the physical affordances of the object.
the 'guidelines' to designing ambient interfaces by gross seems like a helpful rubric for evaluating a design. however, depending on so many factors such as use, intent, and context, many of the standards could be completely subjective. how does one judge utility? or consistency? a lot of the slow technology philosophy runs counter to this 'efficient computing' setup of standards.
globalplug: the swedes are so cool! so are the brits and the dutch and the italians and the japanese! le sigh.

disposables

2005-03-11T15:12:00.000-05:00

new thoughts on ambient project. will write more later, but for now, a glimpse.

+ + + + = news you can use.

stay tuned, folks. it's gonna be good.

submergence thoughts + review

2005-03-11T14:37:00.000-05:00

so i presented the proposal to both my group and to the class. here are some insights:

changed the name from gory to hunkydory. bloodbath has been shed for submergence.
issue with changing either the medium, or the message. general sentiment was that there wasn't a clear side on whether the communication was awareness or actively sent love messages.
rings around the tub to indicate time markers seemed a good idea; could it reflect comparisons? patterns?
what kind of history would be displayed or available? is this daily or tracked over time?
adding an indicator which would reflect a proactive input
water + technology is a cool area to explore
is there a more immediate mapping?
bath-taking might become a time-draining chore
dedication vs. responsibility

other works to check out: natalie jerimijenko's watches and mobile feelings by christa SOMMERER & laurent MIGNONNEAU.

submergence presentation

2005-03-11T14:27:00.000-05:00

did mini-presentation regarding tub project for house_n and for mas961. view the svelte powerpoint show here.

background courtesy of pixelgirl presents.

user-modeling systems

2005-03-06T15:12:00.000-05:00

[papers]

Alfred Kobsa, Generic User Modeling Systems, User Modeling and User-Adapted Interaction, v.11 n.1-2, p.49-63, 2001
read on springerlink

[comments + questions]

i wasnt sure exactly what i was supposed to glean from this paper. i was a little confused in that it didnt quite elaborate on what a user-modeling system actually *did*, rather than what how this system might be used for research or commercial purposes. it was tricky for me to glean the 'why' and 'how' from a fuzzy 'what'. however, it provided a very broad view of its general usage, and brought up a lot of the personalization versus privacy issue. there's a delicate balance between a merchant knowing your personal habits so it may tailor itself to you (for the consumer's benefit + convenience), and advertisers completely occupying your attention with uber-techniqued, scarily specific targeted marketing. i think it might be a way to have completely unbiased (a pipe-dream, perhaps) 3rd parties that mediate between consumer and companies, such as froogle, sidestep, or epinions. imagine a client software that would act as your agent, without feeding you targeted ads in the sidebar! or a consumer could specify the level of privacy / convenience they'd like to leverage.
the paper talks mostly about the advent of web personalization, but i wonder what user-modeling in the physical world would entail. could there be a system which took a time + spatial monitor of an individual's activity and then... what? you could do a parallel with brick+mortar stores. scenario: each time i walk into anthropologie (clothing), i can access all the items i browsed through or tried on the last time i was physically at the store, or some items that i specifically tagged on the online catalog. then perhaps there would be a special rack or dressing room just for me with items that anthropologie thought i might like based on my previous interested purchases.
as to online systems, other than personal accounts at the store, there isn't a great way to identify users, and consequently their purchases, preferences, etc. that comprise the user model. what happens if someone else uses my machine and unwittingly purchased something while i was logged in? it might jumble up the data, or at least put something on my history that's inaccurate. more accuracy --> more invasive identification strategies --> less privacy --> big brother! aah!

project proposal : bloodbath

2005-03-02T00:15:00.000-05:00

What is it & why is it interesting?

the motivation for this project was to connect two people [lovers] who live apart, people who want to relate to each other and communicate more effectively their presence, emotion, and love for the other beyond the limitations of current technology (IM, telephone, webcam, etc). i wanted to incorporate a more tactile, enveloping mode of spanning two remote bodies in a meaningful, tangible, and poetic way.

when apart from someone with which you share an intimate relationship with, many small, yet significant things are felt to be missing: the simple act of touch, a warm breath on the skin, the feeling of a heartbeat. signs which bring a lover to life are the fundamentals, the most human.

the inspiration for 'bloodbath' [i may change this name if people find it a too gory connotation, but the reasons for this will become clearer] is the basic record of a heartbeat. a heartbeat signifies life; changes in the pulse indicate periods of activity and rest, excitement and relaxation. an easy way to 'communicate' this sign is to pulse the signal of one's heartbeat to the other lover in some noticeable way, but i wanted to create a subtle, effective mapping. blinking a light or vibrating something would probably be too distracting, or over-pique the curiosity, or just confuse the mapping completely.

then i thought... the human body is the perfect 'computer', the perfect machine. at rest, the human heart beats roughly 60 times a minute, every hour of every day. that's a lot of pumping, day in and day out! at the very basic level, that's 86400 throbs; add in exercise and other heart-racing enablers, and the number increases. while living life, you don't really fathom the sheer aggregate effort of your heart's efforts.

i want to visualize + embody this fascinating feat, this priceless human energy, into something to be communicated and appreciated fully, preciously, and completely.

Usage Scenario

'bloodbath' takes this element of one's heartbeat, and translates it to a drop of water in the remote lover's bathtub. one tiny drop: 1 mL. the frequency of droplets would directly correspond to the pulsations of the other. droplets one after another wouldn't seem like much, but 86400 mL (minimum amount for a day's worth) translates to roughly 23 gallons of water, which is about half the capacity of a standard bathtub.

at the end of the day, the lover has a full bath drawn for them, and can enjoy and relax in the waters, which were 'shed' from the heart of their beloved. the bather has full control; they can supplement the water or drain it slightly, but the accumulated water in the tub is the embodiment of their lover's liveliness, the physical beating of their heart. to bathe in such waters is akin to enjoying a priceless gift. the tub, sensing that it is being used, sends feedback to the heartbeater, perhaps a warmth on their skin or a glow on their device [more on that in the next section.]

one might argue that the daily bath might take too much time if the other is busy or doesn't feel like a bath. however, this system enforces an implicit commitment to sensually indulge, to take at least a moment of the day to unwind and relax. in long-distance relationships, there's a dangerous tendency to take the other for granted for a day (or two, or three, or...), or obliviously neglect the other for an indeterminate amount [i.e. not calling, not writing]. the act of bathing becomes a metaphor for intimacy, for sharing a special quiet time together, for wrapping oneself in the generous warmth of the other. 'bloodbath' infuses a relationship with something shared that's very ritualistic, very intimate, very enveloping.

[side note: in tom robbin's jitterbug perfume, the bathing rituals of alobar and kudra are one of the key components to their immortality.]

How will it be implemented?

i envision the heartbeat being recorded by a simple heartrate monitor (i'm currently playing with one from polar), a wearable component mounted on a chest-strap. for a female, this situates right underneath the breasts; the monitor could easily be embedded in any sort of lingerie with a bra-like strap around the chest. many innovative, sexy, and pleasurably tactile designs could be developed through this wearable interface. this device would be easy for a girl to wear, since most put on some sort of bra undergarment on a daily basis. for a guy, the heartrate monitor could be on a wristwatch, or possibly embedded within a tight undershirt. in both cases, normal perspiration would lubricate the electrodes.

each heartrate monitor would have a unique address and a unique addressable partner. these pulsation signals would be wirelessly transmitted to a server, which would route and transmit the signals to a receiver in the water valve of the remote tub. netlag wouldn't really be an issue in this system because the transmitted signal is pretty constant... drip drip drip drip.

the tub itself is mostly standard; the user can turn on the faucet or open the drain at will. even if the lover isnt wearing the heartrate monitor for any reason, it wouldn't restrict water flow in the other's bathroom or anything. there's always a way for the user to override the technology in either direction.

however, the tub would be very energy-efficient as far as heat, having some sort of covering, and possessing an effective thermo-insulative property. this way the tub would keep warm, and perhaps radiate heat throughout the bathroom and living quarters, a source of living comfort.

there would be sensors in the tub that would indicate if someone was taking a bath or not. either infrared or weight or light sensors embedded in the bottom of the tub. if it senses that the user is taking the 'bloodbath', it send a signal back to the wearer of the heartrate monitor. the heartbeater would definitely benefit from this feedback (knowing their 'efforts' of love are consummated through the bath], though i'm still musing on the best way to do this. since the device rests on the chest already, something subtly 'felt' would probably be best, either slight warmth or a imperceptible purring.

What parts will you complete for this class?

ideally, i'd like to get the heart rate monitor to control a water valve that's wired up to it, releasing a droplet with each throb, gushing if my heart's racing. i'll experiment with different ways in which to embed the device in a wearable garment... lingerie, belts, wristbands. i will also do some user studies, if only with paper prototypes, to see what long-distance lovers might think of this mode of communication, of a bridging metaphor. is it pleasurable? is it creepy? how much control do they feel of the system? would they be comfortably implementing it in their homes? if i have time, i can experiment with other signals such as breathing, blinking, and touch, and translating that over to... aromatherapeutic infusions? secretive tickling? tiny sparks of spice in every bite?

What do you hope to learn?

i would like to explore the intersection between human social relationships, mapping behaviors via remote delivery, and delving into the spatial architectural interaction which develops. right now there isn't much tactile long-range communications, and i want to find the tricky balance between 'this feels like we're close' and 'this feels like a technology is trying to bring us close'. i want the everyday habits and objects to reflect more of who we are, and how they reflect our values.

ubiquitous computing

2005-03-01T11:16:00.000-05:00

[papers]

Ubiquitous Computing, Weiser 1993
http://www.ubiq.com/hypertext/weiser/UbiCompHotTopics.html

Perspectives article for ACM Interactions, Weiser 1993
http://www.ubiq.com/hypertext/weiser/ACMInteractions2.html

The coming age of calm technology, Weiser & Seely Brown, 1996
http://www.ubiq.com/hypertext/weiser/acmfuture2endnote.htm

[comments + questions]

i'm deeply encouraged by mark weiser's vision of the ideal, information-rich interaction with our environment. i feel as though multiple machines and gadgets (and the pesky maintenance of such objects through power, memory, constant-carrying, etc) get the better of us, and the benefits to the technology are filtered through the clunky interface or division of attention. invisibility is paramount: i'd rather sit in a chair and have it automatically adjust to optimal comfort and configuration than fiddle with a lot of buttons and knobs; i'd like to be able to access what i need to see without having to attend to hundred other displays. periphery + the everyday seem to be the natural solution for our human perception to keep up.

the three obstacles to true ubicomp that weiser ennumerates seem to still be extant: (1) great need for copious wireless bandwidth, (2) dealing with roaming issues and changing addresses for mobile devices, and (3) immobility of current windows systems. even though he writes this in 1993, 12 years later we're still suffering from difficulties. perhaps the situation is not so dire as it was back then, but the infrastructure for the seamless, controllable, always-at-hand technology is still lacking. therein lies the difficulty of HCI development: the user interfaces might be conceptually sound, but implementing them on PDAs or mobile phones or screens-on-random-surface gives a cursory spin on tired systems. what could be a useful way of exploring or researching ubicomp interfaces without being visibly hindered by technological hurdles? are there ways to overcome them by sharing spaces with other ubiquitous things like electricity or writing?

i love his proposal of 'childhood' as the theme for the future of ubicomp. this really takes into account the emotional and sensory pleasure of good design, not merely informational overload or straight-forward tools for the tasks. we should have more delight in the everyday moments, incorporating discovery and pleasure into the necessary and purposeful.

interfaces with common sense

2005-02-23T09:44:00.000-05:00

[papers]

Beating Common Sense into Interactive Applications, Lieberman et. al.
http://web.media.mit.edu/~lieber/Lieberary/Common-Sense/Common-Sense-Intro.html

[comments + questions]

Common-sense reasoning seems to put the everyday human knowledge base into computers, enabling more realistic and rational reasoning + problem-solving. I think this is useful for basic logic-based chaining and more 'unbiased' material, but putting it into the realm of personal communication and more artistic + poetic arenas makes me really uncomfortable. Some examples:

art-, film-, music-makers. The DJ application is completely redundant. The artistry is precisely in choosing "what kind of music particular groups like" and seeing the reaction of the people on the dance floor. It's a subtle skill... a lot more than superficial things are taken into account, like the type of club + location, the hour of the evening, smiles on people's faces, and special circumstances like guests or occasions. Increased/decreased activity isn't the only crucial dancefloor feedback; there's a rhythmic ebb + flow to keep things interesting, unpredictable, human. For the same artistically-blind reasons, I feel that rule-based algorithms (no matter how context-aware or common-sense) detract much depth-plumbing humanity from crafting films, art, literature, etc.
predictive typing for messaging. I have never used the predictive word-completion utility on my phone txtmsg because it never writes what I want! Even if it does "make perfect sense." Maybe it's just me, but I'd rather have slower, fuller control than short spurts of correctness interspersed with an exasperatingly wrong word. Even in the paper, the image features a grossly erroneous grammatic error: "i am am at the train st." Why doesn't the application realize that's senseless?
empathetic email. I don't understand why the affective faces interacting with an email-composer assist the user. Perhaps it might point out newbie mistakes ("all-caps indicates shouting," "avoid emoticons and txtmsg abbreviations in a business message"), but the agent cannot readily detect more sophisticated aspects of language such as sarcasm, poetry, and the nuanced relationship between writer and reader. Modeling point-of-view and personality might help, but it's still difficult to accommodate individual complexities on the level of communication. Amazon recommending a book is one thing; telling me how my best friend might react to my email is another.

Because many tasks are underconstrained and "a little bit of knowledge is better than nothing," I agree that common-sense applications can help steer the user into the right direction (a la funnelcake). However, I'm still very skeptical. AskJeeves.com seemed to invite more casual search queries like "how do I [verb] [object]?", but it still returned results of mixed quality or relevance. I suppose the challenge is to highlight the helpful while hiding the horrible. Somehow we are trained to take a real-life human's words with a grain of salt, but we put incredible trust into machines. Can we build forgiveness into the system?

scenario spectacular

2005-02-23T01:58:00.000-05:00

within the ideal, intelligent world:

The day. I organize my life via central time management scheduler. Events can be added as easily as clicking “remind me about this” on an event email, my saying “I’ll see you next Tuesday at 1” to a friend, jotting a to-do list in my notebook, or seeing a poster outside and running my finger over the details of the event. Every morning (or on demand), a printed-paper version of my day’s events is delivered, in which after a moment’s consideration I can cross out or circle events, which automatically updates the central scheduler. My notebook (with real paper pages to maintain tactile pleasure and allowing spontaneous writings and sketches) has an augmented schedule page, its display dynamically linked and written in e-ink.

Walking to lunch. I say to my companion, “Hey, let’s get lunch. What’s a good sushi restaurant around here? Are there any within 5 minutes walk?” My agent knows that I love spicy food and have a predilection for soft-shell crabs and grilled eel, so it gives a preference for such details and seeks the best in the area. It whispers back, “No, there are none here that fit your lunch budget, and the fish aren’t fresh. However, the red pepper pasta at the place at the corner is delicious. Your friend will probably enjoy the Caesar salad.” I feel a noticeable but comforting tug in my clothing, directing me in the right direction, and then abating. If I need to make a turn, another tug aids me. If I stop to notice a window display or tie my shoes, the tugging yields. I am in control. When the restaurant comes into view, a little ding-a-ling goes off, and I can place my name + reservation on the list from my location on the street. I have the public restaurant-seating chart at my disposal, so I can put dibs on the cozy spot near the back (sort of like an airplane seat reservation). Once I’ve notified my intent, the restaurateurs can view me on their map, approaching the café from the west, and prepares my table with extra hot pepper sauce because they know I’ll ask for more anyway.

The body. I would like better information about my own health. I am able to access information of my dietary intake (calories, fat, fiber, vitamins, etc), body egresses (calories burned, water to be replenished, loss of fat or protein), analyze the patterns, and view real-time analyses to pinpoint if I’m allergic to a certain food, another food causes blood sugar to rapidly rise and fall, and what works for me to eat before a workout, and when. Will I maintain my weight better if I eat at different times of the day? Should I exercise a lot a few times a week or a little bit a lot of the time? Exercise can be dynamically tracked and self-motivated. The machines remember me + my workout history, and suggest the correct reps + intensities like a sensitive, knowing personal trainer. It knows when my ankle’s still recovering from a sprain, so it suggests running alternatives, and advises more helpful stretching and lifting positions to avoid injury.

Public transport. Once I get to the subway station, I can see a real-time map of the trains and estimated amount of wait time, along with what kind of train it is (long, short), if it’s crowded and suggests I get on the next one, which door it advises me to enter (I’ll get out of the train closest to the correct exit turnstile). The station will be expansive, well lit, and a social haven. Not merely a Dunkin’ Donuts to cater to the sleepy commuter, but full-blown cafes, stores, clubs, libraries, lounges. The underground will house a parallel city to the one above, and the liveliness thrives via the circulatory, outreaching power of the subway. Instead of driving to their destination, passengers ride and immerse more deeply into the public + social space.

Driving. Three things I hate most (in descending order): waiting endlessly in traffic, getting lost and disoriented, and being terrorized by offensive drivers. Fortunately, I don’t need to worry about these problems any longer. Cars are automated, so I just climb in, indicate my destination (“to grandma’s house we go!”), and begin the journey. The car knows everything about the roadway in real-time… from the traffic patterns to the construction tie-ups, from lane widenings/narrowings to blocked-off ways, from the cat darting into the street to when it’s time to refuel. The whole traffic system is networked, so my car knows exactly how to park, control its speed, how to avoid pedestrians, and maneuver among the other cars. The system is optimized for speed, so I am assured that I am getting there as quickly (and safely) as possible. I’m free to sit or relax or converse as I please, as if I were a passenger.

Connecting. I want to meet with, talk to, play with, and work with other people. Technology has the power to simultaneously connect (electronic communication), and separate (as an intermediary). Within a social setting (at a party, in a meeting, in a café) I want a personal connection as intimate + analog as slipping a love letter in a pocket, returning someone’s gaze, or complimenting someone on his or her bag. We can exchange signals of interest or personality through mutual interactions with objects in the environment. I put a special bookmark in a magazine and hand it to my friend; they see links to all my favorite articles and images. I clink glasses with the cute boy; he sees a cute message when he finishes his drink at the bottom of his coffee cup. I tap my foot to a certain beat in my head; the jukebox recognizes my tune and cues it up next. We play games on the tables and the walls; my finger traces out a message that disappears like invisible ink and reappears for the next lucky person who ‘decodes’ the message with their touch.

Fashion. Nothing is peripheral; there’s a high level of personal integration. Clothes are an outfit of connected devices which give us more control; they become us. The most intimate and necessary garment, underwear, centralizes the system. It’s safe, close, personal, and worn everyday. Everything else wirelessly communicates within. Tiny speakers in my earrings whisper reminders or broadcasts my favorite song when I’m blue. My sweater traps more air when the temperature drops; thins out to a mesh when it’s warmer. My skirt has a built-in rocket booster for times when I need to turbo-rev my way. Everything I wear is customized; a system scans my body and knows the exact measurements and proportions to make the best-fitting and most comfortable clothes. In all things, we shouldn’t change for technology; we should remain true to ourselves.

class wiki!

2005-02-16T19:16:00.000-05:00

http://wiki.media.mit.edu/bin/view/Ambient/AmbientIntellgenceSpring2005

location-based and context-aware

2005-02-16T10:40:00.000-05:00

[papers]

A survey of Context-aware Mobile Computing Research by Chen & Kotz
http://elans.cse.msu.edu/ni/restrict/ChenKotz2000.pdf

Context-aware computing applications by Want et al
http://www.ubiq.com/want/papers/parctab-wmc-dec94.pdf

[comments + questions]

Especially working among students who are using the sensor-laden house_n placelab for activity recognition and just-in-time information, I am particularly wary of systems which, however adapatable, are still essentially rule-based and vulnerable to unreliability in unpredictable situations. Instead of developing increasingly complex and robust algorithms to interpret high-level human context through low-level indicators (i.e. "she's eating breakfast" from "within an hour of leaving bed, sensors indicate bowl, spoon, milk, and cereal box have all been touched within the span of 5 minutes"), i think real-time displays of low-level, invisible information can be a lot more useful and much more reliable. for instance, a wearable tool that measures UV radiation and warns the wearer if she is prone to sunburn or melanoma. mere knowledge of data to the user, even without suggestions or automated tasks, is very powerful. other examples are energy / water / resource consumption, time stamps / graphed duration of location or disposition, or overlaying a language translation of labels on objects. One thing I wish were invented is a tool that would directly map whatever I eat to my vital statistics (blood sugar, blood pressure) and stress/mood/energy levels (so I could have the knowledge and control to time my meals to optimize my alertness throughout the day, for example). Maybe someone else would want the same tool to tell them that warm honey milk is the best thing to make them sleepy at night.
How much does user-controlled data (directly-input profile, interests, personal details) versus automated data capture (recording trends, following a user's dynamic preferences) influence the system? If they contradict, should the system bias itself a certain way? I can think of times when the user's initial profile would want to dominate, and other's when someone doesn't know themselves very well and puts trust into the system.
Thinking about what humans are good at versus what computers are good at, I reviewed some ways in which humans retrieve our own data through contextual cues. We strongly associate memories with smell, which is a truly biological phenomenon. It might be a silly invention, but instead of a handheld alarm or an augmented display, perhaps the smell of coffee reminds you to remember the keys, or a particular smell is released on the days when you have lab meetings. Taste might even be a trigger, too. Only a subtle sample would be effective; however, such a multifaceted sensory simulator (and nonobtrusive installation and deployment!) might be a long ways off. However, it might be a useful complementary system to our overloaded visual + hand manupulation systems.

intelligence augmentation

2005-02-09T11:47:00.000-05:00

[papers]

Maes, P., Schneiderman, B., "Direct Manipulation vs. Interface Agents: a Debate." Interactions, 4 Number 6, ACM Press, 1997
http://www.cc.gatech.edu/ccg/paper_of_week/shneiderman-maes.pdf

The Wearable Remembrance Agent: A system for augmented memory, Rhodes
http://xenia.media.mit.edu/~rhodes/Papers/wear-ra.html

[comments + questions]

Although a wearable computing device may be hands-free and mobile, how much of an attention and cognitive load would it feed on throughout everyday usage? Is the compromise of intermediated social interaction and heightened distractibility worth the constant at-hand accessibility of relevant information?
I agree with Pattie that the agents are best suited for non-critical tasks, such as personal recommendations and relevant highlights, rather than suggestions for the next step in a highly complex emergency situation. Examples given are websites, books/movies/media, matchmaking, auctioning, and writings. Cues may be ignored or overlooked. However, how formal a situation could an agent behave suitably and helpfully? Suggesting routes via specific trains, schedules, and stations? Strongly enforcing a deadline (e.g. won’t allow you to play a game if you haven’t completed an assignment that’s due in two hours)? Scheduling and prescribing medications?
The two papers didn’t mention the consideration of a multiple-user interface. What if, instead of just connecting two people who shared something in common, it would link entire defined groups such as kerry supporters, yoga buffs, or technology firms? Or, through the wearable device, members could pull up relevant info from someone else in your trusted circle (i.e. mobile google without all the inappropriate or unrelated links). It would be quite interesting to see how communities develop while acquiring and using a collective augmented intelligence.
Shneiderman argues that optimized user performance (“rapid performance, low error rates”) is the primary, if not sole, standard to judge the success of a system or interface. However, I wonder if there are more flexible modes of evaluation that includes variables such as emotional satisfaction, user happiness, and individual connection. The experience of manipulating a program should be efficient as well as enjoyable. Even if a system gets perfect marks for user performance, who would want to use it if it were boring, unresponsive, or impersonal?