In part, this simply reflected the intuitions and inclinations of the designers: good ideas seemed to flow from considerations of boundaries, and seemed to be more difficult to achieve in other areas. To a great degree, however, this move reflects and reveals a new perspective on domestic technologies more generally.

In keeping with the Design Research methods we have been developing, we pursued the theme of Boundary Devices through a series of proposals, low- and high-tech prototypes, culminating in three systems that we are deploying with volunteers for long-term trials.

The term ‘Boundary Devices’ not only reflects our concern to sensitively breach the walls around the home, but to develop designs that bring together several of Equator’s communities. Thus the designs use Equator technologies such as ECT, and take advantage of methodological expertise in ethnography and home studies. Moreover, they pick up on core Equator issues: how to bring together local and distributed information; how to design products that are physically distributed in the home; and how to embody our conceptual insights on issues concerning ambiguity, interpretation, ownership and change.

We describe each of the Boundary Devices in the following sections:

The Video Window

The Plane Tracker

The Local Barometer

The Video Window

The Video Window consists simply of a small video camera mounted outside the house, feeding a live image of the surrounding landscape to a monitor hung inside the home itself. It provides a new view from the home, one that has particular characteristics because of the technologies that are used.

Video Window at Bill’s house.

The Video Window was originally developed by one of the Goldsmiths team members for his own home, not as a research project per se but simply for his family’s enjoyment. The experience has proven remarkably compelling. The monitor is hung in the bedroom, and is often the last thing seen at night, and the first in the morning – because of the characteristics of camera and screen, it often reveals a view when the world appears dark through the physical window next to it. The screen and camera also affect the visual quality of the view: the relatively wide-angle lens offers a wider panorama than that seen through the physical window, and colours and contrast are subtly heightened.

The camera is mounted on a tall pole, so that the scene sways when the wind blows. Combined with visual artefacts of, e.g. raindrops on the screen (which often cause rainbow-tinged distortions), this allows the Video Window to provide relatively direct access to the weather. Nonetheless, the experience is largely an aesthetic one, with the slowly changing lighting creating a backdrop for occasional actions such as the flurry of a bird flying past the camera or the stately swoop of a passing airliner.

The view from the Video Window changes slowly over time.

After the Video Window had been running for more than a year in the team member’s home, the Goldsmiths team decided to recruit a volunteer household to try the system to see whether they would appreciate it in a similar way. Working with the Sussex team, they deployed a modified version of the system in a home in Brighton for more than two months.

The Brighton installation: Camera.

Like the original installation, the Brighton test involved mounting the camera on a very tall (30m) pole outside the volunteer’s household. The video cable was fed into the house near the kitchen, and led to the monitor which sat on a shelf overlooking a table where the household spent much of their time.

One of the significant changes to the original design was in the housing for the video monitor. To distance the display from standard TV’s, video or computer monitors, and to emphasise the filmic nature of the image, the screen is cropped to a 3x1 ratio, similar to that of cinema productions.

Brighton installation: Display.

Ethnographic interviews and observations of the volunteers showed that they enjoyed the experience for many of the same reasons as the Goldsmiths team member. They liked seeing the view over the seaside (which could not be seen from their kitchen, though it could be from the top floors of their home). They enjoyed the visual qualities of the view as well, comparing it to landscape painting and other forms of aesthetic experience. The Video Window was a talking point as well, prompting many questions from visitors and friends. It even seemed to prompt a kind of nostalgic reaction amongst some of the grown children of the family, reminding them of the views they had enjoyed every day from their bedroom windows.

A video clip of one of the participants, talking about his experiences during the trial, may be downloaded here (.wmv file, 1.25 MB).

Our field trial of the Video Window revealed a set of issues for the system as well. It became clear that the volunteers were reluctant to tinker with the system when things went wrong, even if the problem was as simple as the camera pole twisting slightly and shifting the view. In part, this seemed to reflect the complex relationships amongst the design team who ‘authored’ the system, the Sussex researchers who took responsibility for day-to-day maintenance, and the volunteers themselves. But it also reflected the mixed nature of the system itself.

Our understanding of the Video Window has been informed by the ‘Brand’ conceptual framework originally introduced in the Domestic Experience. This framework, based on Stewart Brand’s How Buildings Learn, explores the different patterns of ownership and change that characterise different layers of the home. ‘Stuff’ in the home, for instance – books, bric-a-brac, small appliances – tends to come and go at the whim of the owners. The “fabric” of the home, in contrast – its roof and walls, infrastructure such as plumbing and electricity – tends to change quite slowly, often with the aid of outside professionals, and even under outside constraints such as planning permission. We are not simply in control of everything in our homes: different aspects have different stakeholders, different models of permission and control, and different timescales.

From this perspective, the Video Window is not just a prototype product distributed over two locations (inside and outside), but two different layers of the home. The camera, mounted on a tall pole affixed to the side of the building, relatively difficult to install or service, is close to being part of the fabric of the home. The monitor, in contrast, is more familiar, portable, even replaceable: potentially part of the “stuff” of the home. The mixed nature of the Video Window seemed to account for the way the volunteers addressed (or failed to address) the system. In any case, it adds to the issues involved in creating products consisting of multiple distributed components in the home.

The Plane Tracker

The Plane Tracker uses a purpose-built antenna to pick up transponder signals from nearby aircraft, and decodes information about their origins and destinations. This is used to control the display of high-altitude aerial photography on the device’s main screen, recreating the flights passing by. The intention is to evoke a feeling of visceral connection with the far-off lands linked to the home by the huge metal ships sailing overhead.

The Plane Tracker (form proposal).

The Plane Tracker’s origins dates back to Workbook 1, when a sketch proposed “data compensation” for the noise and pollution caused by passing transport such as planes, trains or motor traffic. A large number of people live next to the paths of such vehicles, suffering the ill effects of their passage without any direct benefits in return. Vehicles might recompense for this by transmitting information, particularly information relevant to their travels.

Compensatory Data proposal from Workbook 1.

Another step in the evolution of the Plane Tracker was taken when one of the design team read a (possibly apocryphal) story of how exotic plants were sometimes found growing under the flight paths of international airports. According to this account, seeds would blow into the landing gear bays and be carried by the aircraft until released when the wheels went down for arrival.

Fascinated by these tales, the designers started experimenting with displaying plants and flowers native to distant airports as a way to indicate passing traffic. When an airplane had been detected, a virtual seed might be planted in the system, slowly growing and blooming over the next several days. As a garden grew, it would give hints about the passing traffic. People might even become accustomed enough to the patterns of virtual greenery that they would notice when unusual events occurred – a new route opening, for instance, or a plane diverted from its normal course.

The Window box would show virtual plants from exotic places.

The Flower Box was a poetic idea, but difficult to achieve to our satisfaction. We built a large database of plants and flowers from around the world, but finding examples that uniquely specified different locations was challenging. In any case, we were uncertain whether people would learn to map plants to locations, and were unwilling to provide too much supporting information (e.g. labels) for fear that this would override the plants themselves.

Using aerial photography for the Plane Tracker grew from our attempts to find alternatives to plants as displays of distant locations. In part, it was encouraged by our discovery of an API for Google Earth, upon which the current system is build. With some support from computer scientists from Nottingham and Lancaster, we extended this to allow us to plot our own trajectories and viewing angles over the aerial photography this provides, and to remove place labels and similar graphic distractions.

In our current design, the display screen is masked to distance it from commercial computer and TV screens (see the Video Window for similar considerations). For this device, we are using a hyper-ellipse (basically an ellipse with four focal points) as the shape for the screen. This is reminiscent of the shape of commercial airliner windows, but we chose it primarily because it works extremely well with the slow arc of the apparent flight and the curve of the horizon that is sometimes visible.

Screen of the Plane Tracker.

Our initial prototypes have used a commercially available aerial for receiving transponder signals. We have been in contact with scientists from NASA AMES, in the US, who ‘grow’ specialised aerials for spacecraft using genetic algorithms. They have generously supplied us with the specifications for several aerials tuned to the requirements of our system. Not only do we expect these to be of high performance, but they embody a strange but motivated aesthetic that emphasises the unconventional nature of this system.

An aerial for the Plane Tracker generated by NASA scientists using genetic algorithms.

Looking at unlabelled aerial photography is compelling but confusing. The system does not approach locations in canonical North-South orientations, and it can be disorienting to view even familiar landmasses from unfamiliar angles. We have experimented with a number of subsidiary displays that might aid orientation. Our current design is for a small physical globe that would be spun on two dimensions to indicate the current view location.

Physical location device (globe).

We have explored a number of approaches to designing the overall form of the system. Alternatives include creating it as a device to be hung on the wall, like an augmented window, as standalone appliance like an alien TV, or as a hybrid of the two.

Form sketches for the Plane Tracker.

The Local Barometer

The Local Barometer seeks to provide people with a sense of the sociocultural texture around their homes: cleaners advertising in newsagents’ windows, nosey neighbours and pub gossip, conversations in Post Office queues, fridges for sale in a local paper.

A wind vane mounted outside the home senses wind speed and direction. Down-wind postcodes are used as search terms to retrieve content from the web, returning classified ads, news items, or images associated with the appropriate areas. This content is transmitted to a network of small displays distributed throughout the home, where it scrolls across the screen in a direction also determined by wind speed and direction. The system suggests the imaginary flow of local information through the home, and works to encourage reflection on the local environment.

Original sketch/collage proposal for Local Barometer.

Inside the home, the Local Barometer consists of multiple small screen displays that are situated throughout the home. Each display has a unique form and is designed to suggest potential locations: sitting on a mantle piece, wedged between books on a shelf, plugged into kitchen sockets, hiding under stairs or hanging by a hook from a door handle.

Form prototype of Local Barometer display.

The displays are small, allowing the user to dip in and out of active engagement with the text and images they show. The output is not intended to serve as a source of information or entertainment, but instead as a series of partial clues about nearby neighbourhoods that we hope will be compelling and provoke curiosity, even voyeurism.

The system has allowed us to explore how to craft text (including found text) to support a poetic form of peripheral awareness. In part we have been inspired by poets such as Walter Carlos Williams in our approach to this.

The format of adverts from classified-ad sites is specific and rigidly formatted, featuring a price, a contact number or email address. These details are removed from the data, leaving just the descriptive body, resulting in a more curious form. The texts become defamiliarised, making the information unclear and creating a certain power and ambiguity. This is emphasised with the addition of a few carefully chosen line breaks.

Poetry Engine.

The line breaks create moments of surprise, when meanings not intended in the original advert suddenly surface as equally valid readings of the text. The less familiar the content appears to be, the more context the readers have to provide themselves in order to make sense of the messages.

The same tactic of decontextualisation is applied to the images displayed on the Local Barometer. Stripped of their associated text, images of objects, people and places float by provoking curiosity and surprise. For instance, is the washing machine for sale or in the wanted section? Is that a picture of a lost kitten? Why is an image of the park on display?

The approach of removing some context from the web-gathered data allows users to produce their own meanings from what they see. All the information displayed on the Local Barometer is rooted in the local environment, but it is not designed as a search tool or a recommender system. Rather, it provides a rich source of content upon which users can layer their own meanings.

Form models for the Local Barometer.