Introduction In this paper, we introduce a new ambient display, personal steganography, and the concept of the urban score. Strictly speaking, the ambient display itself is a particular rendering of a value of the same name. As we will explain, the two are intricately linked. This ambient display does not convey stock prices, bus schedules, remind you to buy milk, or any other such useful bits; it gives the user a feeling--perhaps even just a hint--about their connection to the city they are walking in and its other inhabitants.
We argue that the display shown in Figure 1 is both an efficient display of a great deal of information and is well designed for its ambient task. Its task is to convey, likely helpful information to the user in a way that is both unobtrusive and always present. What cannot be depicted in Figure 1 is that this display is not typically shown at full brightness and it is always at the lowest level of the window stack or “in the background” on your PC. It is a resource that the user can draw on when they “aren't doing something else” or when they are between other actions. We in no way mean to criticize the display, its many brethren, or its authors.
Rather, we want to argue that there is another task of interest that, while it shares some constraints with Google's widgets in Figure 1, opens up a different and important design territory. This task is more closely related to exploring a new city or “neighborhood, walking into a restaurant or bar to "see what it's like”, or chatting with a friend about local political events. The task is feeling the pulse of a city. The idea of the urban score is that somehow measures and conveys that pulse.
Any mention of an urban “score” quickly leads to
a discussion of the rulesets used to calculate
such a number (or set of values) and this
discussion often is followed by more vehement
arguments about which set of inputs or outputs more
validly describes one person's urbanity versus
another. We encourage this argument and seek
to foment it. We would like to see many
designers come up with their own metrics of urbanity
and have users compare, use, and find those
metrics that suit their tastes; we want to
encourage people explore what it means to be urban
through these low-intensity displays. In the next
two section we will discuss two categories of
designs that we are proposing for an urban score.
It seems clear from recent work in industry and
academia [
One urban score design avenue that we have
been considering is taken from the world of
nuclear engineering and radiology, the dosimeter.
The idea is to make an ambient display for a
mobile device such a phone or Ultra-Mobile-PC
(UMPC) that measures your "dose" of the city.
This display could be a background or a screen
saver in the simplest case. In any case it should
be unobtrusive and require little, if any, of the
user's active attention or input.
In our first variant of the dosimeter, we measure
airborne pollution that the user is exposed to.
Many pollutants, such as carbon monoxide, sulfur
dioxide, nitrogen dioxide, ozone, and fine
particulate matter can now be measured with cheap,
handheld instruments [
This simple urban score is a display that shows the total and highest amounts of pollutants a user is exposed to over the course of the day. This could be displayed as simply a mixture of background colors, with no text at all, making it very low-demand in attention terms. Although there are certainly communities who would find this type of ambient information both useful and important, it fails, in our view, to spark a debate about how one is experiencing the city. This type of design might even be better as a more direct, non-ambient, display by compiling information from many users into standard maps showing the geographic relationship to exposure levels. A second design that is superior in our view, is the dosimeter that measures the amount of a city's "vibe" as your urban score and then gives a more ambiguous ambient display we called a personal steganographic ambient display. The name comes from steganography which is the art and science of writing hidden messages in such a way that no one apart from the intended recipient knows of the existence of the message; this is in contrast to cryptography, where the existence of the message itself is not disguised, but the content is obscured. A typical steganographic application is to high messages within images via alter low order bits of pixels, etc. In our approach we alter small portions of what appears to be a regular image. The alterations are subtle and occur over long periods of time making them hardly perceptible to the casual untrained glancer. However, to the person who knows how to read the display, a wealth of information is stored within it. The image shown in figure 3 is one example of a personal steganographic display. It shows a view of the skyline of Shanghai, PRC and has many easy to manipulate dimensions. For this example, the figure shows only two of the possible ways to modulate this display, the height of the tower at the left and the number of smaller buildings shown. These and many other properties could be easily layered within the image. In our effort to promote conversations, even arguments, about life in the city, we claim that it is better to entangle the notions of the display, the measurements taken, and the mapping between them. This encourages exploration and opens up new avenues of dialogue. We claim that there are any number of metrics that one could measure about the city, the user, or other people that could be used as fodder by designers. We will demonstrate here a simple, two-dimensional, easy-toimplement display that could be generated in real time on a mobile device at any time.
For this urban score, we map the height of the
tower to the number of other people that you have
encountered over some interval. Are you "out on
the town" or "stuck at home?" This can be
measured easily with bluetooth scanning. Even though
only a small fraction of people turn on their
bluetooth radio and make it visible for scanning, we
can safely assume that this proportion is roughly
constant so that seeing twice as many bluetooth
phones indicates roughly the twice the number of
people. Naturally, this part of the urban score can
be manipulated to include well-known persons, be
they friends [
The second dimension, shown as number of
secondary buildings in Figure 3, a running average of
your proximity to the city's "center" for some
spatial definition. The latitude and longitude could be
easily measured by GPS, now common on mobile
devices, or by some approximation based on
visible beacons [
An effect of this display is that commuters who live out of town would see the city unfurl or disappear as they went through their day. The skyline would remain "distant" on the weekends if they do not venture into the city for non-work activities. A slight variant of the measurement computed here would be to take the geographic centroid of a set of friends' movements and measure from this point. If there is a "standard hangout" for the gang, it would become the center of the city for that group of people and the display would change accordingly.
In Figure 4 we demonstrate two envisioned personal steganographic view of an urban score for San Francisco, USA. The views exaggerate a wide range of changes. In reality, a very small handful of changes would be subtly occurring at any given time. For example, the sky color change could indicate air quality with the smoke from building representing sulfur dioxide (SO2) specifically. The crane could indicate the arrival of new buildings as you approach your “center” of the city. Strangers and familiar strangers are captured by the birds - both flocking and perched. There are also balloons, airplanes, boats, flowers, and tents whose number and appearance can map to specific elements of an urban score. Similarly with animals (grazing) and people (walking, picnicking, and sunbathing.
An alternate approach is shown in Figure 5 using a more iconic representation of various elements to “render” an urban score. In this example as you move from suburb to city the screen “slides” to reveal the urban image (top). The sky color shows air quality comparing air quality where you are now with sensors in the city. Familiar strangers are represented by people and unfamiliar strangers by birds.
We would argue that a key factor that differentiates cities, especially major ones, from towns or the rural environment is leadership, "being ahead of the curve." It is hard to imagine fashion trends coming "to" New York, London, Los Angeles, Paris, or Milan. Similarly, musical trends often coalesce around cities' both large densities of musicians and people eager for new experiences: Hip-hop in New York city, R&B and soul in Detroit, downtempo in London and Manchester, and hairmetal on Los Angeles' Sunset Strip. Films are routinely shown only or make their debuts in major cities, to say nothing of the film festivals. Part of living in a city is the stimulation, even excitement, brought by new things happening in front of your eyes.
For the Cloud type of visualization, again we seek to score what is latent in the experience of a city. A naive Cloud visualization, such as the one shown in Figure 6, would be to take the tags of the songs that people near you are listening to on their mobile devices and form a tag cloud. These tags are usually categories or the names of artists, but experience with last.fm (music) and flickr.com (images) shows that users will tag in useful and unexpected ways. User contribution also implies wrong or distasteful contributed tags; note the misspelled tag in Figure 6 "electroic" is more popular than say "chill." Also the tag "czilaut kompletny" (really "chillout completely") is a joke on (easing of? mistranslation of?) slavic or eastern european languages that represent a significant fraction of the listeners to this type of music. The relative size in Figure 6 indicates the number of times one of the authors listened to songs with that tag, however in an urban score it would be more interesting to map size to the preferences or current selections of those people nearby. This could be easily sensed among people with Apple iPhones, Microsoft Zunes, or bluetooth enabled devices that share music information. In Figure 4, the position of words is not a controlled dimension, it is alphabetical. Of course this dimension could be easily controlled as well, in the simplest case putting words radially closer to the center if that tag was sensed "recently." With this measurement and visualization, a user walking down 5th Avenue could get a sense of what the world is listening to and why they live "in the city."
We dubbed the visualization shown in Figure 6
naive because it can really show very few of the
properties that make cities trendsetting. It is, or
soon will be, possible to do this for music, as
shown, and perhaps photography as cameras
emerge with networking capabilities. This is
insufficient in our view to get a view of the
complexities of the trends in the world's cities.
We propose a new scheme based on credit card
sales transactions. In its simplest form, the music
tags in the figure would be exchanged for goods/
services recently purchased by those nearby or
perhaps the names of stores they patronized.
This would present a much more accurate portrait
of activities in the city and all the data is collected
already (by credit card companies) and is already
available to most users on the internet. Almost
needless to say, though, this presents a privacy
problem of the highest order. If this were actually
the desired design it is likely that a technical
scheme could be devised to "hide" a single user's
data amongst the great multitude, ala mix routers
[
In this section we will give a brief overview of
some of the evaluations of an urban score that we
think might be interesting research contributions
to the community.
• Measuring the different "uses" that a mobile
device takes on. This evaluation would
compare a personal steganographic urban score
visualization with a more functional one, such
as in Figure 1. Within a subject, it seems clear
that news and weather have value, but how
does that value (both by usage and perception)
differ from the value of the urban score?
When? How does this tie in with the idea of
promoting “wonderment” in cities [
In this paper we have introduced the notion of an urban score, partly a low-attention display and partly the basis on which that display rests. This "score" allows users to put their finger on the pulse of a city, to get a feeling of those around them, and question assumptions about their urban life. We have introduced the idea of a personal stegonographic ambient display. We have suggested a variety of mostly personal stegonographic visualizations of the urban score to kickoff the debate about what an urban score should measure and how it should be visualized. Two of these displays, the dosimeter category, attempt to show you how much of the city you have consumed, via inhalation or inebriation. The last two focus on understanding those that are near you a common situation in densely populated areas. These two cloud visualizations could be generated from easy to sense values and provide insights into a community. With these designs have try to highlight our significant concern for designing usable, thought-provoking systems that protect the users privacy.
We hope that these thoughts can be a springboard to others.
What's your urban score? What would you like it be?