Liz Goodman did a nice taxonomy of networked ojbects in her midterm paper for this class in spring 03. She laid out her taxonomy according the conditions required for their function, or the interactions they enabled, as follows:

• "Mobility: Does the functionality of the device/Technology gain anything from the user's location in the physical world?
  
• Reciprocity: Does the service/device enable one-way or two-way communication? How does it function in a social network?
  
• Relation to network: what is the device's place in the virtual network?"
  

From this, she divided the projects we'd looked at into four rough categories:

• Paired Objects: "which function as surrogates for a user in a remote location (cf. Greenberg & Kazuoka). They enable reciprocal communication between two or more in separate locations. Typically, they are tied to a location or locations associated with the users."
  
• Network Displays are location-independent. Communication is typically one-way; network displays pull data from a network, either disembodied information like network traffic, or physical information that's been distilled into digital form, such as weather data.
  
• Remote Controls "use physical proximity to deliver location-specific information and services. As such, they're usually mobile." One service they can deliver is the ability to affect other networked objects, in which case they also function as paired devices (cf home alarm systems, Alzado.net)
  
• Community Devices for example, the Lovegety. "These devices combine location-based functions with autonomously communication between devices", enabling communication between parties who may nor may not have known they have something about which to communicate. In this, "they act as semi-independent agents, not surrogates."
  

Liz cites community devices as the most interesting and least investigated class of networked objects, and she's probably right, though there has been a rush of interest in them in the last year or so.

Liz mentions that she doesn't see these as mutually exclusive categories. I think they're better thought of as endpoints to a pair of axes that lay out the possibilities. One axis runs from paired objects to community objects, and another from network displays to remote controls. Moving along the paired-community access, there's an increase in connectedness and a greater need for meaage connections to be made quickly and broken quickly .

On the display-remote control axis, there's an increase in the importance of both reciprocity and a tight interactive loop, and a decrease in exposure. In other words, fewer people at a time can see the remote control at any given instant, and a tighter response time is needed between the control and the thing controlled.

...

Mobility has been a hot topic the last few years, spurred by the advances made in wireless networking. While wireless makes for some exciting possibilities, there is a great deal to be explored in the realm of what I'll call architectural coincidence (someone's gotta have a better term for it). Built environments incorporate a great deal of thinking about the traffic flows of people. In well-designed buildings and social spaces, this thinking pays off, and people follow regular traffic patterns through these spaces. In doing so, they leave traces of their presence on the space. These traces are not often collected, but when they are, they can provide useful information about the people who use the space, how the space can be used, and more.

For example: every time I go to the gym, I swipe my ID card to go through the gates. So do many other people. The result, if collected, can be looked at in a number of ways. We can look at the habits of individuals over time, or we can look at the patterns of groups moving through the space. We can see when traffic is highest, when it's lowest, we can sort the traffic by people's relation to the space (e.g. students, faculty, staff, etc), and more. This kind of information is of obvious use to building managers, architects, and administrators, and is easily collected and shared using a network.

However, there are other possibilities to consider. For example, whenever I go to the gym, I'm usually going in between appointments or classes, and I have the next appointment or class on my mind. I like to use swimming time to think about what's next. What if I used the network and the building's awareness of me as a chance to remind myself of upcoming events? What if I allowed other people to do the same? This is similar to the idea of location-based messaging that many researchers and designers have considered, with a slight difference: it relies on the location, not a device on my person, to deliver the message.

Paul Berry had another project idea which took advantage of architectural coincidence to do some social engineering. The "How's your day?" project consisted of a series of three buttons, mounted in control panels in non-work areas of a workspace (water cooler, bathroom, etc) and a display asking a question. The question, "How's your day?" could be answered by pressing the "Great" button, the "Okay" button, or the "Lousy" button. The display would then show you the aggregate of everyone's responses. Paul's idea was that this could potentially shift the mood of the office one way or another by letting everyone see the patterns of their co-worker's moods. In class, we discussed how different locations might affect the responses, as some people might not want to push "Lousy" in a very public space; how placement near the boss' office or the conference room might skew the answers; and so forth.

...

Posted by tigoe at September 2, 2003 08:37 AM