Meghan, an 85-year-old woman whose memory is failing, sits in a comfortable chair in the living room of her home in Dublin, Ireland.
On her lap rests what appears to be an oversized pincushion with a small circular biofeedback device embedded. Two wires protrude from the device, and Velcro loops with embedded sensors are attached to the end of each wire. Meghan slips the loops over her fingers, flips a switch, and a small screen flickers to life. A horizontal, wavy line appears and spreads across the screen; after 30 seconds, the line stabilizes. Meghan presses a grey button on the device. She takes a deep breath and utters the word “attention.” The graph spikes upward, signalling an increase in alertness.
Meghan (not her real name) is participating in an alertness training programme designed by the Technology Research for Independent Living (TRIL) Centre —a groundbreaking research collaboration, funded by GE Healthcare, Intel Corporation and the Irish government, to address the physical, cognitive and social consequences of ageing. The goal of the TRIL initiative is to develop new technologies to help older people maintain their independence and to age in place from wherever they call home. Researchers from three Irish universities and Intel participate in the TRIL Centre, applying a multidisciplinary research model and actively involving study participants in co-creating technology solutions. The evolution of the alertness training programme illustrates the benefits of this unique approach to programme and technology design.
Promoting independence
The alertness training programme arose from research showing that memory and other complex mental functions are strongly influenced by a person’s state of alertness. Attention and alertness are particularly important for older people, influencing how well they function at home and in their communities, and affecting their ability to live independently. Lapses in attention can lead to accidents or falls—and potentially to institutionalization and loss of independence.
Alertness levels can decline in older people, due to decreased activity and declining levels of key chemical messengers in the brain, such as norepinephrine. The training programme is designed to help older adults learn to boost their alertness levels temporarily, ‘on demand,’ so they can keep track of their goals, get things done, avoid time-consuming disorganization and costly errors, and even prevent accidents. The ultimate goal is to help older adults lead more focused, engaged lives.
The TRIL researchers set out to design techniques and technology that older people could incorporate into their daily lives in order to monitor and improve their alertness levels. The self-alerting technology would leverage biofeedback from the electrical conductance of skin, more commonly known as galvanic skin response (GSR); research has demonstrated the potential for biofeedback training to address attention deficits.
Designing the programme
The design initiative was launched in 2008, when a team of TRIL Centre neuropsychologists, led by Ian Robertson, Professor of Psychology at Trinity College Dublin, began to lay the groundwork for an in-home trial of a self-alerting technique (SAT) and technology. As a first step, background studies were conducted in Robertson’s lab to determine whether older people could learn to modulate their alertness levels. Roughly 40 participants participated in the lab experiments, which confirmed the viability of the self-alerting training concept.
With the feasibility of the training concept demonstrated, the team recruited seven older adults to participate in an exploratory research phase. Participants were recruited from a group of older people who had been assessed at the TRIL Clinic in St. James’s Hospital in Dublin and identified as having difficulties related to alertness.
The seven recruits were put through a five-day training programme administered in their homes by the researchers. The programme included pre- and post-training cognitive assessments, to determine the impact of the training. Participants went through two in-home training sessions, using a laptop and an off-the-shelf GSR biofeedback device wired to the computer. They also were asked to use the SAT daily and record their experience in a notebook. They were shown how to log onto the Internet (access was provided as part of the study) so they could participate in three remote telephone sessions with a researcher later in the week.
Design challenges
This preliminary research identified a number of design challenges that had to be addressed: It was cumbersome for researchers to carry equipment and supplies to participants’ homes. The in-home training and cognitive assessments were time-consuming for researchers and participants. The week-long process also was intrusive for participants, who had to allow researchers into their homes for extended periods of time.
There were technical challenges as well. The GSR device was unreliable; it was often difficult to find a signal. And despite the fact that engineers worked to make the technology as simple as possible, it was still too difficult for some of the older people to use the laptop and GSR device, and the limited training the researchers delivered proved insufficient. “Older Irish adults are not used to technology,” says Agnieszka Milewski-Lopez, a senior research assistant in Robertson’s lab who represents the neuropsychology discipline on the core team. “Many of them still don’t use mobile phones or computers. So we learned that it was important to build technology that would be easy to use and non-intrusive.”
Despite the challenges, the initial experiment confirmed the usefulness of the SAT concept. Some participants reported that they found the technique helpful and were starting to use it in their daily lives. The next step was to redesign the study so that it was less dependent on researchers and less intrusive for participants.
A key aspect of the redesign would be the creation of a more user-friendly technology solution, with the help of an interaction designer and ethnographer who joined the core research team, and co-principal investigator from Intel, Muki Hansteen-Izora, who came on board to work with Robertson. The goal was to ensure that the training would incorporate not just the knowledge of neuropsychologists about how the brain works but also would reflect the needs and preferences of the older adults who will use the self-alerting technique and technology.
The three researchers on the core team worked together to ensure that the insights of each of the disciplines they represented—neuropsychology, anthropology (with a focus on ethnography) and design—would be integrated to create a successful design solution. “What makes our research so effective is our close collaboration, and not pulling the study in the direction of one discipline or another,” says Milewski-Lopez. “We determined that it was important to find the areas of commonality across our disciplines and focus on that as we designed a solution, challenging our ideas and not letting our professional biases distract us from our mutual goals.”
Participatory design: encouraging engagement
The new core team adopted an iterative, participatory design approach that involved close collaboration with a new group of eight older people recruited for this phase of the research. The concept of participatory design is based on the notion that inviting people to co-create technology ensures that the ultimate solution will reflect their needs, and as a result, they will be motivated to use it. It’s a powerful concept, and one that was critical to the TRIL researchers, as their goal was to develop a training programme that older people would self-administer in their homes, without the aid or prompting of a researcher.
This “bottom up” approach is a departure from the typical technology design process, which focuses more on the engineering aspects of design and less on users’ needs. It is a key differentiator of the TRIL Centre’s work. “Brain training is such a huge area of research,” says Robertson. “There’s a lot of awareness of the need for older people to stay mentally alert. But there’s been less emphasis on compliance, on ensuring that people will actually use their training over time. We wanted to give older people a tool they would embed in their daily lives, because they helped to design it, and therefore it is meaningful to them.”
Designing the “alertness cushion”
Eight older adults were recruited for the next phase of the research, carried out in 2009. Working closely with the new participants, who had never been exposed to alertness techniques, the researchers redesigned the GSR biofeedback device from the ground up. The result looked nothing like a piece of technology.
The new device is a circular, cushion-like design roughly 30 centimetres (approximately 12 inches) in diameter—a sort of oversized pincushion. The “alertness cushion,” highlighted in the opening anecdote, is designed to be very simple to use. When the device is switched on, it takes a baseline GSR reading. The user presses the grey button before self-alerting, and the on-screen graph spikes upward to indicate a higher level of alertness, as measured by galvanic skin response. The GSR data is recorded on a micro-SD card built into the device, for later analysis. The finger sensors leverage the SHIMMER sensing platform, developed at Intel and now licensed to Realtime Technologies.
What evolved into a cushion began as a box. When Flip van den Berg, the interaction designer, joined the research effort, he recommended that the team strip away any features of the GSR device that were not absolutely necessary to conduct the study—a process he calls “undesigning.”
Evolution of the “alertness cushion” (left to right): Researchers began with a biofeedback device in the form of a simple black box. With the active participation of older adults in the alertness training programme research study, the box evolved into the form of cushion with an embedded biofeedback device. Velcro loops with embedded sensors, based on Intel’s Shimmer technology, capture galvanic skin response (GSR). The GSR data is recorded on a micro-SD card built into the device.
So the team designed a GSR device in the form of a small black box with sensing wires attached. The idea was to present older people a “blank slate” and let them help to design a device that would match their needs. “By using only a black box and thereby not hinting at any specific form factors, participants could focus on how they would like the device to look and feel, so we could create something meaningful and useful for them,” says van den Berg.
The researchers observed participants using the black box and how their usage changed as they became more comfortable with it. They probed to learn what features the older people wanted the device to include and what form it should take. The eventual cushion design was inspired by observing one participant rest the box on a cushion on her lap. The cushion design makes the device appear less medical or technical and therefore more approachable to non-technically inclined older adults.
The in-home pilot of the training programme
In addition to redesigning the biofeedback device, in preparation for the home deployment, the researchers modified the training content based on what they’d learned from participants. For instance, they decided to introduce the SAT and biofeedback device later in the training, after participants had time to learn about alertness and observe their own alertness levels, so the new tools would be more useful and relevant to them. They also found that participants wanted to use their own alertness words as they performed the SAT; doing so made the training more meaningful to them and helped to build engagement. The use of personal alertness words would be added to the pilot programme. Finally, the research confirmed that older people valued the flexibility of the training, which could be performed with a mobile device any time, anywhere, making it easy to embed in their daily lives.
The pilot was implemented in 2010. Twenty older people were recruited to participate. Each received an alertness training kit, which includes a GSR biofeedback device (the “alertness cushion”), a guidebook and two audio CDs—a tutorial on self-alerting and a CD focused on self-alerting practice.
Building awareness of alertness levels (Week 1)
In the first week of the programme, participants learn about the concept of alertness and begin to build awareness of their own alertness levels, and how they fluctuate. They are asked to reflect on activities that make them feel more or less alert, and how this affects their lives (the guidebook poses questions to guide their reflections). They are presented with a list of 26 words related to alertness (e.g., focus, attention, aware, sharp, alive) and asked to choose the words that resonate with them. From this list, they choose their own personal alertness word, which they will use later, as part of the self-alerting technique that will be introduced in the second week of the programme.
Over the course of the week, participants identify areas of their lives where they might use the alertness training (e.g. to improve memory, energy, motivation and concentration). They also commit to practicing the SAT, and they choose specific days and times to practice.
The self-alerting technique (Week 2) 
In week two, participants learn more about the impact of alertness on memory, energy, mood and other aspects of their lives. The self-alerting technique (SAT) is introduced and participants are asked to begin practicing it twice a day, and whenever they feel the need. The brief audio tutorial (about eight minutes long) guides them through the technique, which involves a five-step process of sitting upright, breathing in and out slowly, saying their alertness word, taking a few normal breaths, and noticing how they feel. After performing the SAT, they are asked to reflect on the experience, recording their thoughts in the guidebook.
During the week, participants are given examples of how to put the SAT into practice (e.g., for remembering names, or where they left their car keys, or to stay alert when performing everyday tasks such as cooking, cleaning and driving). They are asked to set simple goals for applying the SAT daily, using their own alertness word to personalize the process and motivate them to employ the technique.
On the last day of the week, a photo of the alertness cushion is presented and participants learn that the device will help them to monitor their body’s response to SAT, by measuring skin conductance levels. The goal is to familiarize participants with the new technology they will learn in the following week, without overwhelming them.
Biofeedback training and the self-alerting technique (Week 3)
Week three focuses on biofeedback training using the alertness cushion. Participants learn how biofeedback works and how the cushion device can help them to gain more control over their alertness levels. A detailed instruction manual, with plenty of photos, guides participants through the operation of the device. They learn how to interpret the biofeedback line, and to be aware of the factors that make the line move—body temperature, movement and posture, pressure on the finger sensors, mood and thoughts, and breathing. By themselves, the movements are meaningless, so the older people must learn to read the graphs in context, to gain insight into what is making their alertness levels increase (or not) at any given moment.
After the first day of using the device, participants complete a questionnaire that asks whether they understood why they’re using the biofeedback device and if they feel confident in using it and interpreting the biofeedback line. If they respond “no” to any of the questions, they’re encouraged to call a researcher whose phone number is prominently listed. “We didn’t want people who didn’t feel confident using the device to decide not to use it anymore and drop out of the programme,” says Milewski-Lopez.
The researchers received only a handful of calls for assistance; they were expecting many more. “The feedback that I’m getting is that most people were really happy with themselves on day two because they understood how to use the alertness cushion. This really helped to build up their engagement for using the device.”
A focused life: the beginning (Week 4)
In the final week of the training programme, participants focus on applying the self-alerting training in their daily lives, with or without the biofeedback device. They are given a goal diary to complete for the week, and asked to note whether using the SAT helps them to be more alert when performing three activities of their choosing.
No new concepts are introduced this week. Rather, the older adults are asked to reflect on what they’ve learned in the programme and to continue practicing the SAT. The programme ends, but the researchers hope that this will mark the beginning of a more focused life for participants.
Scaling to meet the needs of thousands
Researchers are in the process of gathering data from the GSR devices and guidebooks, and are preparing to conduct post- training cognitive assessments. To reduce the burden on researchers, the assessments will be done via telephone rather than in person. The process will be partially automated, leveraging technology developed for the TRIL Centre’s “Dear Diary” research project, which is devising a technique to conduct remote assessments of cognitive function via telephone interviews. A follow-up telephone interview will be conducted one month after the end of the programme, to determine how successful participants have been in applying the alertness training.
Once the results of the pilot are analyzed, researchers will refine the alertness training programme, incorporating the input of the older people who have helped to co-design the training at every step of the way. “I am struck by how much people have written in their guidebooks,” says Robertson. “It’s quite extraordinary, and we want to capture this rich information in the next version of the programme.”
Researchers plan to conduct a randomized clinical trial of the revised programme in the second half of 2010. In the meantime, they are enthusiastic about the response to their research thus far. Many participants in the home pilot said they plan to continue using the SAT in their daily lives, and several report they are doing so already, in a variety of settings—when driving, playing golf (just before putting), during lectures and concerts, when playing bridge, doing household chores, and conversing with friends and family. “I’m really excited about how participants have engaged, and how important this programme has been to many of them,” says Milewski-Lopez.
Robertson believes the TRIL Centre’s unique, people-centred approach to design has been critical to the success of the research effort. “By involving many disciplines and working closely with older adults to identify their needs and perceptions, we were able to create a training programme that truly engages and motivates older people,” he says.
The continuation of the research beyond 2010 will require ongoing funding of the TRIL Centre. “We are very enthusiastic to see this research continue in 2011, and we’re hopeful that TRIL will receive funding to carry on this important work,” says Robertson. “We’ve designed a cost-effective programme that people can self-administer in the comfort of their own homes. There is tremendous potential to scale this approach to improve the quality of life for thousands of older adults. That’s a very gratifying thought.”