Designing for seniors

XIV.4 July + August 2007
Page: 38
Digital Citation

Exploring the nuances of Murphy’s Law—-long-term deployments of pervasive technology into the homes of older adults

Janna Kimel, Jay Lundell

There is an increasing focus in industry and academia on developing technology to allow older people to remain in their homes for as long as possible [1, 6]. One approach involves "pervasive technology"—the deployment of various sensors, computers, and output devices embedded in the home environment to track behavior and interactions with household objects (Figure 1). Behind the technology sits software that measures, for example, activities of daily living (ADLs), accumulates this information, and uses it to notify relatives or medical people of the status of an older person living alone [5]. Additionally, this software could provide direct assistance to the older person, such as activity-appropriate prompts and reminders [2].

Since 2002 a group of researchers at Intel’s Digital Health Group have been observing and interviewing senior citizens in studies to better understand them, how this "ubiquitous" sensing might help them, and determine design issues to address. Our philosophy has been to develop emerging-technology solutions and place them into "real" homes as soon as possible, rather than use an on-campus simulated home. This approach provides more valid data on the "livability" of our solutions.

Initially, we conducted extensive ethnographic interviews of the home environment and the challenges faced by older adults [4]. Next, we deployed sensors in the homes of six senior citizens and the homes of their primary caregivers to better understand their social activities and how feedback might help improve their social contact [3]. Our next project has involved sensing daily activities and medication compliance to deliver "context sensitive" medication reminders. We have enrolled 11 older adults in a long-term study that is just now being completed. Here, we summarize many of the practical considerations in deploying this type of technology and working with older adults in the design of in-home technology.

Diversity in Older People

There is a tendency in our society to think of adults over 65 as a relatively homogenous group. In fact, older adults are extremely diverse. As people age, they begin to experience differential declines in a variety of sensory and cognitive systems. Some older people experience early hearing problems, while others have great difficulty with their vision. They can have fine-motor problems, mobility difficulties, memory and cognitive problems. Older people can exhibit any combination of these issues, from so mild that they seem perfectly healthy to those severe enough to require medical care.

Older people also exhibit great variety in their rate of decline. We have worked with people age 85 and older who are healthier, more active, and physically younger than others who are 65. Thus, segmenting even by age groups alone is not a reliable method. There are great differences in social connectedness, orientation to technology, income level, interests, and health status. Thus, in designing technology to accommodate this wide variety of elders, designers must be very creative and technology must be very flexible and adaptable to the changes older people experience.

The Importance of Technology Disclosure

Many older adults have little or no experience with computers or sensors. Thus, the first hurdle in our trials was explaining each sensor and its purpose in the home. Some sensors were more intimidating than others. The motion sensors placed in each room have a red light that blinks when activated (Figure 2). For some, that red light looked like the flash indicator on a camera. From this, we have learned a good rule of thumb: demonstrate each device and show any data output gathered from the device (Figure 3). Of course, all of the safeguards of data protection and privacy must be clearly explained, and the participant must fully understand the purpose of the study and give informed consent to participate.

Issues in the Placement and Validation of the Pervasive Technology

Prior to installation, each home was carefully mapped out so that sensors would provide reliable and unambiguous data (Figure 4).

Initially, the team had planned a "classic experimental" approach to our study in which the sensors would be placed in the homes of the participants at the start of the study; there would be little contact with the participants during the study and finally a post-study interview and data analysis. We quickly realized this was completely unrealistic. Often our engineers or researchers would need to visit the home, and there was much interaction with the participants during the study. There were a variety of reasons for this, as explained below.

Unanticipated data. In our initial interview with participants, we spoke with them extensively to assess their patterns of behavior so that we could anticipate what the sensors might show. Without knowing a participant’s typical behavior, it is often very difficult to interpret the sensor data or even determine if the sensors are working properly. For example, for one participant who lived alone, the bed-sensor data often showed unusual "spiking" patterns each morning. Upon reviewing her interview data, we recalled that she exercises in bed each morning, and we realized the spiking was her daily exercises, not a defect in our sensor.

In spite of frequent contact, participants would often leave unexpectedly for a few days. In these cases, it was difficult to know whether something was wrong with our system or if the participant was gone. It was surprising to us how often even quite frail elders would leave for vacation or an extended stay in a family member’s home. In spite of our requests that participants notify us when they would be gone, they often forgot to tell us.

In our most recent study, we placed a seven-day-type pillbox fitted with sensors in the homes of older people and told them to take a vitamin from the pillbox twice a day. We expected to see that people who were compliant would open the pillbox door and then close it twice each day. Instead, we saw every imaginable variation. Some would take out both pills in the morning and leave the nighttime pill on top of the box to remind them. Some would leave the doors open so they could see if they took the medication and then close them all at the end of the week when they refilled the boxes. In their eagerness to maintain the spirit of what was asked of them (take their vitamin twice daily), our participants had actually been providing very puzzling information. Our assumption about behavior was incorrect, but this also led to potentially misleading data since pill taking was not captured unless the door to the pillbox was opened.

Accidental mishaps with the technology. Although we tried to place sensors where they would not be disturbed, there were many accidental mishaps (Figure 5). Participants would rearrange their furniture and accidentally knock a sensor off the wall. Often, this went unnoticed until we observed no activity from that sensor for a few days. Some homes had housekeepers who would come in and disturb the sensors. Another participant apparently thought a sensor needed cleaning, as she placed it in the dishwasher! One participant had a dog who found one of our sensors, took it out to the backyard, and chewed it up.

Again, to avoid as many accidents as possible, a full explanation of the sensor, what it does, how to care for it, and the reason for its placement in a particular location may have alleviated some of these issues. Our initial intent was to unobtrusively place the sensors about and have the participants ignore them. We now know that it’s best to educate the participants about these new devices in their homes.

Participant interventions with the technology. Some participants intentionally moved the sensors. Said one woman, "I didn’t like where that motion sensor was—I wanted to put a picture there." We noticed another sensor was frequently not working, only to find out that the participant often placed her jigsaw puzzle box top directly in front of the sensor while working on her puzzle.

We used a bed sensor that was placed underneath the mattress to detect patterns of sleep. In one case, we concluded that the bed was not being used, as no data was coming from the sensor. When we visited the home, we found that the participant had put it into her closet because she thought it was a misplaced heating pad.

In our first study, we had designed a laptop application that we wanted participants to use. Knowing that some participants might be intimidated by computers, we placed a plastic cover over the keyboard and fastened it with Velcro and placed a large trackball on top, so that all that was required was the use of the trackball (Figure 6). However, we discovered that several participants would occasionally remove the keyboard cover, cancel the experimental application, and use the laptop to play solitaire.

Lessons Learned

We are currently analyzing eight months of data from 11 elders in our current study, in which we have learned a great deal about deploying pervasive technology into homes. Real life is chaotic, even for single elders living alone. Older people go on unanticipated vacations, or illness can strike, requiring extended hospital stays. Pets, grandchildren, and housekeepers can wreak havoc with your system.

Older adults are willing and patient participants when you explain the importance of the research, even for extended studies that require significant commitment. One participant, after several months in the study, asked, "Will this technology help me?" We answered, "Maybe not, but the research may help many people in the future." She smiled and said, "If I can contribute to making things better, I’ll be in your study as long as you want."

Elders are not necessarily afraid of technology. The onus continues to be on the designers, who should actively engage elders in testing new technology and elicit their thoughts and opinions.

Best practices for deploying pervasive technology for older adults:

  • Test your technology extensively in a controlled lab setting, and then be prepared for it to not work anything like it did in the lab.
  • Older people can learn and use technology and computers—if you design it properly and provide a reason for them to use it.
  • Leave contact information—participants will need to contact you when things go wrong.
  • Leave detailed explanations/instructions of the technology and how to interact with it.
  • Call often to check up on them and keep them motivated. Phone calls also provide a forum for unanticipated issues the participants may not think to call the research team about.
  • Plan to have plenty of support people to react quickly and fix problems.
  • Plan to have daily meetings with the entire team to review the data and determine if there are problems that need to be fixed. Whenever possible, include the elder’s extended family. We found that the best participants often had a tech-savvy family member such as a son or daughter. This helped to keep the participant motivated and comfortable.

Doing technology fieldwork with older adults requires much care in preparation, and the research team needs to expect the unexpected. Murphy’s Law is the guiding principle in this domain! But remember, older adults are willing and engaged participants and should not be overlooked for valuable input.


1. Dishman E, J.T. Matthews, and J. Dunbar-Jacob. "Everyday health: Technology for adaptive aging." in Technology for Adaptive Aging. Edited by Hemel RPSV. The National Academies Press, 2004, 179-204.

2. Mihailidis A, L. Tse, L., and A. Rawicz. "A context-aware medication reminding system: Preliminary design and development." in Rehabilitation Engineering and Assistive Technology Society of North America. Atlanta: 2003.

3. Morris, M. "Social networks as health feedback displays." IEEE Internet Computing (Sept-Oct 2005):29-37.

4. Morris, M., J.Lundell, E. Dishman, and B. Needham." New perspectives on ubiquitous computing from ethnographic study of elders with cognitive decline." Proceedings of Ubicomp 2003: Ubiquitous Computing 2003:227-242.

5. Rowan J, and E. Mynatt. "Digital Family Portrait Field Trial: Support for Aging in Place." Proceedings of ACM CHI 2005 Conference on Human Factors in Computing Systems 1 (2005): 521-530.

6. Stefanov DH, Z. Bien, and W. Bang. "The Smart House for Older Persons and Persons with Physical Disabilities: Structure, Technology Arrangements, and Perspectives." IEEE Transactions On Neural Systems And Rehabilitation Engineering 12 (2004), 228-250.


Janna Kimel
Intel Research

Jay Lundell
Intel Research

About the Authors

Janna Kimel began her professional career in theatrical design and transitioned to medical product design. After creating costumes to delight and inspire an audience, Janna ran a business for 15 years designing and manufacturing apparel for elders and people with special needs under the label Accessible Threads. Her past design work also includes consulting with innovative companies—such as IDEO, Herbst Lazar Bell, and SC Johnson—to create medical and household products. Since joining Intel’s Digital Health Group as a design researcher in 2005, Janna has had the opportunity to focus on the needs of elders while both conducting research and designing health products for the marketplace. As part of the User Experience Design team in health, research and innovation, she works with elders to more effectively design.

Jay Lundell received his doctorate in cognitive psychology in 1988 from the University of Washington, where he studied decision making, expert knowledge, and computational theories of cognition. His industry research has focused on human computer interaction for in-home consumer products. Jay’s current interests are in assistive technology for a variety of health issues—cognitive impairment; daily activities and social engagement for older adults; and exercise, diet, and medication adherence. He is also interested in the efficacy of health-related user interfaces, such as "just in time" prompting, goal reminding, and historical trending. Jay is currently working on medication adherence for older adults, using intelligent tracking of daily activities to provide appropriate, context-aware reminding for medications.


F1Figure 1. Pervasive in-home technology for older adults

F2Figure 2. Motion Sensor

F3Figure 3. A typical graph of the type of sensor data we were collecting. The top figure tracks motion, and the bottom figure tracks when a person is in/out of bed.

F4Figure 4. Floor plans had to be accurately drawn

F5Figure 5. The refrigerator sensor was hard to place and was often knocked off

F6Figure 6. A laptop with a keyboard cover and large trackball

©2007 ACM  1072-5220/07/0700  $5.00

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

The Digital Library is published by the Association for Computing Machinery. Copyright © 2007 ACM, Inc.


Post Comment

No Comments Found