Embedded Technology (Ubiquitous Computing)

Do walls have ears? Not right now, but it will not be long before walls not only have ears, but will also be able to see what we are doing and even tell us things that are relevant to our activities. Traditionally, when people said that walls have ears, they suspected that someone was spying. In the modern context, the walls will be a lot friendlier. They will sense who is in their vicinity, but only to determine the occupants' needs and to help them by adjusting the room light and the window shades to suit their tastes or by performing similar tasks.

Such service is an illustration of embedded technology (or ubiquitous computing) at work. In simple terms, ubiquitous computing allows computing architecture to be embedded into the environment. Artifacts in the environment can then sense different aspects of their surroundings as well as the user's ongoing activities, reason about them, and act accordingly.

Types of Ubiquitous Computing

Ubiquitous computing can take various forms. A sampling of some of those forms follows.

Portable Computing.

Laptops and handheld computers have made computing portable. You can carry your computer with you everywhere, but your experience is only slightly different than in your office—you still must interact with the computer through a more or less traditional interface .

Pervasive Computing.

Smart devices have computing technology in unexpected places. At first this will be in information appliances such as phones, personal digital assistants (PDAs), and pagers. Later, pervasive computing is expected to expand to include more traditional appliances such as toasters, refrigerators, washing machines, ovens, home security systems, and so on. Still later, the infrastructure will develop so that smart devices will include equipment like you might find on the road, in an automobile, in a hotel, or in an airport. For example, you might be able to pay tolls or buy airline tickets with your phone-based electronic wallet. Prototypes demonstrating such capabilities already exist, but their widespread usage will require a universal infrastructure with increased computing and communications power.

Smart buildings are buildings that are well networked and equipped with smart appliances and have a personal computer based system that can control the ambient conditions within the building. The "smarts" are needed to start and stop different appliances to maximize the users' comfort or security and to minimize cost. For example, a smart house could run the water heater whenever energy prices are low; it could select the best times to turn on the heat or air conditioning within the premises, based on which rooms are currently occupied or are likely to be occupied shortly.

Calm Computing.

This technology carries out the idea of computing fading into the background. With this technology, the artifacts are intelligent, but do not require focused interaction—we just live with them and they work unobtrusively to make our lives simpler. An example of this is the "dangling string" network monitor. A traditional user interface for monitoring computer network traffic would capture a lot of data and try to present it on a computer screen. By contrast, the dangling string hangs from a hallway ceiling and is controlled by a small motor. Network activity causes the motor to kick the string ever so slightly. Activity in the network is thus presented as the literal hum of the swaying string.

The basic concept behind calm computing is to develop user interfaces that are not based on symbols. Although symbols can convey a lot of detailed information, people have to focus on them to extract that information. In contrast, humans can capture a lot of non-symbolic information in a manner that does not require their explicit attention. Although your main current task can be successfully carried out in a symbolic mode, you can be peripherally aware of a lot of other activities without concentrating on them. The hum of a string, the shadows cast by a ceiling fan, the reflections cast by a shimmering pool of water all fall into this latter category. You will probably be aware of the flickering shadows and reflections and how fast they are changing even though you might be concentrating on your word processor or a phone conversation. Now imagine if these shadows and reflections and other changes in environment were used to convey specific meanings such as "All's well in the plant" or "There seems to be a lot of relevant activity in one of the chat rooms I visit" or "Suddenly, the stock market is seeing a lot of trades in the technology sector."

Wearable Computing.

In this opposite of calm computing, instead of having computers embedded in the environment, you carry the computers on your person. One main difference from regular computers is in the user interface. Wearable computers are designed for hands-free operation. Often, the intended user would be walking about or in an awkward posture, such as atop an electric utility pole or inside a narrow submarine engine room. Equipment repair was one of the earliest applications for wearable computing. It allows the user to give commands by voice and view information through a head-mounted display that projects an image on to the user's glasses.

The earliest wearable computers were quite bulky: the user literally had to carry a backpack with a computer in it. However, with advances in technology, it is possible to have wearable computers embedded in wristwatches or pendants or even sewn into clothing. These miniature computers are usually designed for specific applications, such as displaying text and pictures and giving online directions as the user walks about. In another application the wearable computer provides an interface to a three-dimensional information space where the user's head, neck, and eye movements can be interpreted as desires to probe some aspect of the space in more detail. Instead of traditional virtual reality (VR) , the user may be engaged in the "real" reality and only occasionally looking into the information space when a specific need requires it.

Related Technology

The techniques involved in ubiquitous computing are as diverse as its applications. Many of these techniques, however, deal with peripheral technologies. We need motors to drive the changes in environment that calm computing requires. We need head-mounted displays for certain kinds of wearable interfaces. For the wearables, we need lightweight processors that consume low power and small batteries that can feed them. Certain kinds of pervasive applications—for instance, those fixed in a tollbooth or a smart home—may have no restrictions on size, power, or communication ability; for applications that entail mobility, all those restrictions apply. For wearable computers, there are significant challenges in materials to weave conducting channels into fabrics. Some researchers are developing techniques to generate power from the normal actions of a human body, such as the impact of the heel on the ground in walking.

Interestingly, although the basic programming required to build ubiquitous applications is the same as programming for other applications, the abstractions involved tend to be quite different. The interface modalities of desktop computing are not much use in calm and wearable settings. Another special feature of calm computing is that its effectiveness relies on a lot of reasoning to infer the user's desires. Unlike desktop computing, users do not simply type in or use their mouse to indicate their commands—the system has to figure out if users would like the light to be turned down ever so slightly to adjust to their mood. Thus artificial intelligence (AI) techniques will help here.

Ubiquitous computing tends to require a significant infrastructure. Depending on the needs of a specific application, the infrastructure should include the ability for different components to communicate on wireline or wireless networks; the components should be able to discover each other's presence as do the Java-based Jini services and other registration services; the components should be able to move around physically while retaining their identity and address as in mobile IP and 3G (third generation) wireless networks. An application might require the ability to authenticate participants through the public key infrastructure (PKI) , or the ability to make secure payments through SET, the secure electronic transaction protocol. The exact requirements from the infrastructure will vary with the application, but broad-based capabilities are likely to be needed in many cases. For example, buildings should have location sensors so that they can detect users; users' wearable computers should be able to talk to the buildings and to the computers of other users. Computers in automobiles should be able to talk to computers at tollbooths and be able to pay for the privilege of going through the booth.

Most of the technologies necessary for ubiquitous computing exist and the infrastructure is spreading more widely every moment. Although it is still unknown exactly what forms ubiquitous computing will take when it becomes a commercial reality, we can be sure it will be something both challenging and creative.

see also Ergonomics; Microchip; Operating Systems; User Interfaces.

Munindar P. Singh

Bibliography

Ishii, Hiroshi, Sandia Ren, and Phil Frei. "Pinwheels: Visualizing Information Flow in an Architectural Space." In Proceedings of the Conference on Human Factors in Computing Systems, Seattle, Washington, March 31–April 5, 2001.

Schilit, Bill N., Norman I. Adams, Rich Gold, Karin Petersen, David Goldberg, John R. Ellis, and Mark Weiser. "An Overview of the PARCTAB Ubiquitous Computing Experiment. Roy Want." IEEE Personal Communications 2, no. 6 (1995): 28-43.

Shenck, Nathan S., and Joseph A. Paradiso. "Energy Scavenging with Shoe-Mounted Piezoelectrics." IEEE Micro 21, no. 3, May/June 2001.