This dilemma is all too common as the Baby Boom generation ages and people live longer.
But imagine a network of sensors and computers installed throughout your mother's home. The system could warn her of impending problems, remind her of important routines, encourage her to get some exercise and even call emergency services, if it could not get her to respond.
Scary or exciting as this type of decision-making computer network may sound to you, it's probably going to be a reality early in this century. Its applications are far reaching and seemingly only limited by human ingenuity.
Computers with human-like perception will emerge as society breaks away from the traditional desktop computer and moves into the era of ubiquitous computing. Also known as ubicomp, it refers to an environment where computers are constantly present, seamlessly integrated and applied for everyday uses.
"The traditional desktop computer helps to an extent with daily life," says Dr. Irfan Essa, an assistant professor in the Georgia Institute of Technology College of Computing. "But we believe the computer should not be limited to the desktop. It should be a part of the room."
If computers are constantly present, then they should also be smarter, according to researchers in Georgia Tech's Future Computing Environments (FCE) group.
"The next generation of technology will have computers understanding what people are doing and what they want," says Associate Professor and FCE Co-founder Dr. Chris Atkeson. "The basic expectation is that humans have a clue. For now, computers don't have a clue."
Since 1995, the FCE group has been investigating the ubicomp and aware computing concepts in "living laboratories." These are technology-rich Georgia Tech classrooms, offices and now an experimental house called the
Broadband Institute Residential Laboratory. The researchers, who make up the FCE research team, refer to their work there as the "Aware Home" project.
photo by Gary Meek
|
|
Researchers created a prototype "aware" room to test sensor technologies that make a computer's perception more human-like.
(300-dpi JPEG version - 525k)
|
|
When the Residential Laboratory opens early this year near the Georgia Tech campus, the house will be capable of knowing information about itself and the whereabouts and activities of its inhabitants.
The laboratory, under the guidance of The Broadband Institute, was funded by a $700,000 grant from the Georgia Research Alliance. The 5,040-square-foot house will host a broad range of communications-related research funded by a consortium of more than 20 information technology companies, says Broadband Institute Director Dr. Nikil Jayant.
The three-story Residential Laboratory includes: a basement with a high-performance computing and shared home entertainment section; and two independent, two-bedroom living areas. One living area will serve mainly for experimental purposes; the other will host actual residents, initially students and eventually an elderly person or family.
"Aware Home" researchers want to build an environment that can sense the inhabitants by seeing, hearing and measuring contact through a variety of sensing technologies, including video, audio, motion and load.
"We will be breaking new ground with the Aware Home," Atkeson says. "The computer will be aware of who people are and what they are doing, rather than needing a human being in charge of the remote control, for example. This is the next generation of computing."
With the expertise of seven full-time faculty members, the FCE group is working as a team "to demonstrate how ubiquitous computing can impact favorably and possibly negatively," says Assistant Professor and FCE Co-founder Dr. Gregory Abowd. "We want to understand those different impacts of technology. Aware Home is our next big frontier."
Broadband Institute
Residential Laboratory
The Broadband Institute Residential Laboratory was founded to conduct research on the confluence of broadband communications and lifestyle computing, and their potential impact on quality of life.
In addition to the Aware Home project, other research in the Residential Laboratory will focus on its use as a "connected home," with all forms of broadband communications into and in the home. In fact, the aware capabilities in the residential lab will depend on communications connectivity, explains Broadband Institute Director Dr. Nikil Jayant.
The Residential Laboratory will serve as a core research facility for a variety of Georgia Tech groups, including the School of Electrical and Computer Engineering (ECE), the Information Security Center and the School of Textile and Fiber Engineering. For example, textile and ECE researchers may collaborate on adding wireless connectivity to a wearable motherboard designed by textiles Professor Dr. Sundaresan Jayaraman. Then senior adults wearing the motherboard could send out signals from sensors monitoring their medical condition. Other examples of their work include cross-collegiate experiments in wireless multimedia, ubiquitous computing and private telemedicine in the home.
A consortium of more than 20 information technology companies will fund research in the Residential Laboratory. Among those companies are BellSouth, Eastman Kodak, EchoStar, Intel Corp., AT & T Corp., Nortel, Broadcom, Convergence.Com, Cox Communications, Digital Furnace, Ericsson, General Instruments, Lucent, Motorola and Sprint. "The consortium is interested in the technology as well as the applications of it," Jayant says. "Most of our sponsors have a fairly broad outlook on research to be conducted in the Residential Laboratory, and its long-term impact."
— Jane M. Sanders
|
|
Researchers in the FCE group are simultaneously focusing on human- and technology-centered studies in the Aware Home.
"The human challenge with this technology is as much a challenge, if not more than the technological challenge," Abowd says.
Researchers want to prevent information overload, avoid invasion of the occupant's privacy and create practical ubicomp applications for the everyday user. They have determined the most important potential users initially are senior adults.
An Aware Home initiative called "Aging in Place" is aimed at finding ubicomp technology applications that will allow senior adults to live independently in their homes as long as possible. The benefits are both social and financial.
Eventually, ubicomp technology in the home might be less costly that the $2,000 or more per month it can cost to live in an assisted care or nursing home facility, says Assistant Professor Dr. Beth Mynatt.
Specifically, "Aging in Place" would program the Aware Home to: sense and identify potential crises, and then automatically contact services as needed; augment a senior adult's memory; and track behavioral trends by creating social connections between senior adults and their relatives.
In terms of crisis intervention, basic sensing technology could help relatives determine when an incident has occurred or prevent it from occurring. For example, the Aware Home could alert the resident when the home is getting dangerously cold. It could ask, "Are you doing this on purpose?" Researchers, including Dr. Wendy Rogers in the Georgia Tech School of Psychology, are addressing how to effectively communicate with occupants.
Another goal of "Aging in Place" is memory augmentation, or cognitive support, which helps people in their day-to-day routines. For example, senior adults often deal with the difficult problem of interruption. If senior adults are preparing a meal and get interrupted by a knock at the front door, they sometimes need help remembering what they were doing when they return to the kitchen. The aware system would jog their memory by offering displays of key snapshots taken by vision sensors in the kitchen before the interruption.
The third objective, behavioral trend tracking, is what Mynatt calls "the peace of mind quotient." She and her students created "Digital Family Portraits" for family members to follow their senior relatives' routines and activities, both daily and over time. It also gives senior adults insight into their relatives' lives.
photo by Gary Meek
|
|
Sensors in the furniture and in the windows are among the "Aware Home" components. They will help residents and researchers monitor activities in the home.
(300-dpi JPEG version - 547k)
|
|
The frame of the picture, which would be a flat panel display, is dynamic. Age-appropriate, engaging icons in the frame can give relatives a sense of how the senior adult is doing, Mynatt explains. The icons represent notions of health, relationships, activity and events. Using sensing technology, the Aware Home can get a general sense of whether the senior adult interacted with a lot of people today, for example.
The three bands in the frame represent different periods of time. The center band represents today, the second band represents a summary of the past four days and the third represents the past two weeks. Icons in the band decrease in size from the center to the outer bands to represent the various time periods. The icons vary in density to represent quantity in each of the four categories. Then family members could recognize, for example, that their senior relative's activity level seems to be going down over time. That might prompt them to investigate further.
"They won't necessarily be diagnosing the problem; it's just that this sort of contact makes the situation seem less scary," Mynatt says.
In the technological arena, FCE researchers are studying how ubiquitous sensing can give computers a decision-making context, like humans have. In the Aware Home, Essa, Associate Professor Dr. Aaron Bobick and other researchers are building a context-aware development infrastructure they have tested in a controlled situation in an office.
"Imagine a computer that knows you are near it, knows you are looking at it, and knows who you are and what you are trying to do," Essa says. "Such abilities in a computer are hard to imagine, unless it has an ability to perceive people."
In the Residential Laboratory's Aware Home, researchers hope to make computers perceive things and interact with the user. Various types of sensors capture information on the user, including their location, facial expressions and gestures. The computer might be able to determine, for example, if the user is angry or in pain. The intent, Essa adds, is to make the sensing transparent to the user and in no way create a burden.
Meanwhile, Atkeson and other FCE researchers want to build fundamental models of human behavior to train computers in decision making.
photo by Gary Meek
|
|
Ten strategically sized and located, force-sensitive load tiles (shown here) will be installed in the Residential Laboratory to gather footstep data from occupants. Called the "Smart Floor," the system is a natural input device that can identify and locate a person based solely on his or her footsteps.
(300-dpi JPEG
version - 619k)
|
|
"Can we start to learn the preferences and model the behavior of people in the house by watching what they do?" Atkeson asks. For example, during the workweek when they wake up early, the computer system would turn on the coffee maker. On the weekend, when they sleep later, the coffeepot would be programmed to start later.
The Aware Home's context-aware infrastructure will also interact with wearable computers the occupants may use. Researchers led by Assistant Professor Dr. Thad Starner, a widely known authority on wearables, will explore that interaction.
Another technology-centered investigation at the Residential Laboratory will involve a system called the "Smart Floor," a natural input device that can identify and locate a person based solely on his or her footsteps. Ten strategically sized and located, force-sensitive load tiles will be installed in the home to gather footstep data from occupants. Researchers believe the system can correctly identify the user more than 90 percent of the time. The system's applications are in trend tracking, crisis intervention and security, among others.
Finding lost objects is yet another tracking and sensing technology researchers will study in the Residential Laboratory. The system will use small radio-frequency tags attached to various objects (keys, wallets, glasses and remote controls) the user wants to track and a long-range indoor positioning system to track these objects. The user will interact with the system via LCD touch panels placed strategically throughout the house. The system will guide the user to the lost object using spatialized audio cues (e.g., "Your keys are in the bedroom.").
Researchers acknowledge the system isn't foolproof. Another person may walk off with the keys, for example. In these cases, other tracking technologies, such as the "Smart Floor," can assist in locating the objects. For instance, if the keys were last seen with Sue at the front door at 8 a.m., the system can inform the user of this fact. The user can then conclude that Sue accidentally took the keys with her to work.
All of this technology must be seamlessly integrated, and that software challenge faces Assistant Professor Dr. Blair MacIntyre. "Figuring out how to use different displays for what purposes is a problem that interests me," he says.
center photo: ©1999 Photodisc
|
|
Digital Family Portraits created by Dr. Beth Mynatt, an assistant professor of computing, help family members follow senior relatives. The picture frame, which would be a flat panel display, is dynamic, using icons to give a sense of how the senior adult is doing.
(300-dpi JPEG version - 454k)
|
|
If the future is ubicomp, how long will it take to get there? Estimates vary, but FCE researchers believe houses equipped with Aware Home technology could be available within a decade. With ads now for "Smart Home" devices, consumers might wonder what more they might need in 10 years. But Aware Home technology is really different.
"The critical difference is the current technology has people telling computers what to do," Atkeson says. "The next generation of technology will have computers understanding what people are doing and what they want."
But researchers have a difficult task facing them.
"There's a lot of technology development ahead of us to be able to do these things," Essa says. "But it's exciting to have the Aware Home as another avenue to explore computational perception. The house takes this research to the real world."
Jayant, who's in the position of overseeing research in the Residential Laboratory is determined to take full advantage of this "real world" experimental environment. "We want to make sure that two to three years from now we have measurable results, that we really did important work here," Jayant says. "... Our goal is to find the most applicable technologies for the everyday user."
For more information, you may contact Dr. Gregory Abowd, College of Computing, Georgia Institute of Technology, Atlanta, GA 30332-0280. (Telephone: 404-894-7512) (E-mail:
gregory.abowd@cc.gatech.edu)
