What Matters: August 2002
Carebots: The Future of Home Healthcare
By Colin Angle '89, SM '91 and Amanda Gruber '86
The beginning of the age of practical robots will not be ushered in by a robotic child knocking on your door 20 years from now. It has already begun. Robots have been making their way into our homes, cars, and workplaces for years. Home robots make bread and coffee for us. They help clean our dishes and clothing, and soon they will be cleaning our floors and mowing our lawns. As robotic appliances improve, our homes will increasingly maintain themselves. By performing routine tasks for us, these robots provide us with more free time to engage in other activities.
Perhaps it is time for robotics to focus on being a part of a solution to one of the major crises of our time—caring for our aging population. The technology required to address the needs of the elderly has been developed more recently, and is primed to continue to grow if we choose to invest in it. Although the need for a realistic plan to care for the increasing number of elderly is not new, it is becoming more critical as our population continues to age and time continues to pass without such a plan.
We as a society have neither the financial commitment nor the trained medical personnel required to care for the growing number of elderly. Unless we embrace a new paradigm, that of providing health care for the elderly at home, we will be increasingly unable to provide quality care to the elderly. Not only is it less expensive to care for an elderly person at home, most elderly people report that they prefer to remain living in their homes for as long as possible. Thus, robotic technology that allows elderly people to remain in their homes longer can be a key part of this new paradigm.
Many people's gut reaction to the suggestion that can robots can provide care for elderly people living at home is that robots cannot replace people. And of course, this is true. The robot we envision would facilitate and enhance interactions between elderly people and their caretakers, not replace them.
The key technology underlying near-term eldercare robots enables people to control a mobile robot over a secure connection from any internet browser. This telepresence capability allows a remote person to "physically" visit and interact with someone, without the expense of traveling. These internet-connected robots can be thought of as "physical avatars" into which any one with a web browser and the permission of the robot's owner can project themselves.
From a technical perspective, a physical avatar robot is an engineering marvel. The robot must be intelligent enough to safely navigate around someone's home on its own, because the internet connection between the remote operator and the robot may be lost or interrupted at anytime. And even with a reliable, high bandwidth connection, the amount of data the robot can send the operator about its environment is limited. The web browser interface used to control the robot must be so intuitive that little or no training is needed, so that the operator can focus on the social interaction, not the robot. And the robot must be designed so that people feel comfortable with it, and do not need to learn any new skills to have gratifying interactions with it. For this type of social experience to happen, the robot must respect the basic rules of human interaction, such as respecting another's interpersonal space, and project basic gestural cues used in human interactions, such as those which designate where the robot's attention is.
iRobot has created several of these robots. Although there is clearly more work to be done, the capabilities of these robots are very exciting. iRobot's first commercially available internet-connected robot is named CoWorker. Detailed information about CoWorker can be found on the web.
Now that internet-connected robots exist, what can they do to contribute to maintaining elderly people in their homes? Let us start by looking at reasons why elderly people end up having to move into assisted living facilities or nursing homes. More than half of the elderly population suffer from one or more chronic illnesses for which they have been prescribed three or more medications. Non-compliance with medication regimens is one of the most common reasons for relapse and progression of chronic illnesses, and costs over $150 billion per year in the U.S. alone. In fact fewer than 10% of those people on 3 or more medications comply with the prescribed dosing regimen. Typically this non-compliance is not due to patient resistance, but to the challenge of remembering how to take the medications properly. There have been many attempts to solve this challenge without resorting to robotics, but nothing has yet been done which comes close to having a nurse come find you and dispense your medications to you.
Two other common reasons for physical and cognitive decline among the elderly are social isolation and lack of physical and mental activity. Family, friends, and community members could log onto the robots and spend quality time with their elderly relatives or acquaintances. Family who are unable to visit very often, due to distance or job demands, would be able to virtually visit by logging onto the robot. Community members who are unable to drive themselves, but have access to a computer with an internet browser, would be able to maintain relationships with elderly friends who are home-bound due to chronic illnesses. Physical therapists could log on to the robots and continue to work with patients on rehabilitation programs after they have been discharged. Ideally, all elderly people would be doing some type of physical therapy, because it decreases falls and increases physical and mental well-being.
A further benefit would arise if the elderly person or a caretaker became concerned about their state of health. Today when you call the hospital a nurse will interview you to determine if you need to come in or not. Without any visual feedback, this screening process is not terribly effective, and many people are needlessly rushed to the emergency room, with obvious financial consequences.
Okay, so why a robot? An often asked question whether web cams, a computer, and a stationary medication dispenser couldn't perform these tasks less expensively than a mobile robot. The answer is that a mobile robot provides a friendlier, more effective and more intuitive interface for both users. It also maintains the dignity of the robot's owner by protecting their privacy.
From the elderly person's perspective, the interface of the non-robot solution involves significant invasion of privacy. Many people say that they would like to install web cams in the homes of their elderly parents, primarily so they can locate them if they don't answer the phone, but also so they can monitor their medication or help them find misplaced items. However, these same people say they would never want web cams in their homes because they would feel like they were always potentially being watched. A mobile robot solves this privacy issue by being an entity that is either with you or not with you, so follows the normal culturally accepted rules about privacy in one's home. If the robot is with you, it is just like having a guest over. If it is not with you, then you have your privacy. From the caregiver's perspective, the embodied interface is much more intuitive and natural.
As labor costs continue to rise, the complexity of medical treatments and drug regimens continue to increase, and the number of people in need of such treatment also continues to increase, our ability to provide quality medical care to people, especially elderly people, is in jeopardy. Making matters worse, many areas of the country are experiencing nursing shortages. These shortages are likely to increase because enrollment in nursing schools is dropping, the average age of existing nurses is increasing, and the proportion of younger potential caregivers to elderly people in need of care is decreasing. Our current partial solution of importing nurses from less developed countries is becoming less feasible as these countries are developing nursing shortages of their own. Although home robots will not be able to replace nurses and other health care personnel, they will help those people to leverage their time. In addition, home healthcare robots will facilitate family, friends, and community members who would like to contribute to the care of elderly people.
Sound good? In the time it has taken you to read this article you could have done a crossword puzzle with your grandmother or helped her find the pepper grinder.
About the Authors
Colin Angle earned SB and SM degrees in electrical engineering and computer science from MIT in 1989 and 1991, respectively. During his time at MIT, he worked at the Artificial Intelligence Laboratory, where he developed Genghis, the first of the famous MIT "insect" robots, and later Attila, a walking robot with 18 degrees of freedom. In 1991, he co-founded, with Helen Greiner '89, SM '90 and MIT's Prof. Rodney Brooks P'10, a company to develop intelligent autonomous robots for laboratory use. That company, originally called Artificial Creatures, Inc., was later merged into iRobot Corporation, where he now serves as CEO. Based in Somerville, Massachusetts, iRobot develops a wide range of robots that function as toys, industrial helpers, and military technologies.
Amanda Gruber earned an SB degree from MIT in 1986 and an MD from Tufts University in 1990, with a specialization in psychiatry. She currently serves as an Assistant Professor of Psychiatry at the Harvard Medical School as well as a clinical psychiatrist at McLean Hospital in Belmont, Massachusetts. Her research interests include affective disorders; cognitive effects of long-term marijuana and anabolic-androgenic steroid abuse; and use of computers in psychiatry.
What Matters is a guest opinion column written by a different MIT alumnus or alumna. The views expressed are entirely those of the author and do not necessarily represent the views of the Alumni Association or MIT. Interested in writing a column? Email whatmatters@mit.edu.
