What Do Users Want From “Smart” Wheelchairs?

Abstract

Technological advances have made it possible to integrate a variety of functions into “smart” wheelchairs that include robotics but also sensors that support telehealth and communication applications. The purpose of this study was to identify possible features that current older adult wheelchair users identify as being most helpful and most important and to use scenarios to determine acceptability of potential features. A small pilot study including 5 wheelchair users residing an assisted living facility was conducted that included a questionnaire and a focus group. The participants identified safety-related features as most helpful and important. Within the scenarios presented, the safety-related features were identified as highly acceptable. Features that were available with current technology (e.g. communication with family) were least highly rated. While more research is needed on user preferences, “smart” wheelchair developers will want to include user preferences and focus on safety and telehealth features.

Introduction

Technological advances in wheelchairs, robotics, telehealth systems and other biomedical systems have made it possible to integrate stand alone technology into one device. The integrated device is commonly referred to as a “Smart” wheelchair. “Smart” wheelchairs have the potential to passively obtain advanced telehealth information and also to perform surveillance functions for falls and injuries that have not possible in the past. While the benefits of these advanced features are clear in the early or prompt identification of illness or injury, there is also the potential that these advanced features may be seen as intrusive by users.1

Smart wheelchair technology is in the design phase at present2–3 and the next step is to determine what actual wheelchair users would find helpful in a “smart” wheelchair, what they would find intrusive or bothersome and what features they would recommend to the “smart” wheelchair design team. The purpose of this pilot study was to identify possible features that current wheelchair users identify as being most helpful and most important and to use scenarios to determine acceptability of potential features or “apps.”

Background

Despite their widespread use, wheelchairs do not meet the needs of a substantial proportion of users. Based on a survey of 200 clinicians, approximately 9–10% of wheelchair users find it extremely difficult or impossible to use a wheelchair for necessary daily activities and 40% of wheelchair users report difficulty or impossibility with many steering and maneuvering tasks.4 Furthermore, these 200 clinicians asserted that nearly half of their patients were unable to control a power wheelchair by conventional methods and would benefit from an automated navigation system. This population includes, but is not limited to, wheelchair users with impaired sight, those with hemispatial neglect/paralysis as well as those with significant cognitive impairment. Additionally, individuals who experience spasticity or tremors have reported difficulty using the user-interface components of conventional or power wheelchairs.

A gap in this body of research is that much of the research on wheelchairs has been conducted with younger age groups (i.e. children with physical disabilities, those with spinal cord injury or amputations). The quality of research on powered mobility devices for middle-aged and older people has been described as “very low” (p. 674) using internationally recognized criteria5 with none of the reviewed studies including robotic wheelchairs. With the projected growth in the proportion of older Americans and concerns about increasing physical disability based on rates of chronic disease among middle aged Americans (i.e. obesity, diabetes), the need for powered wheelchairs among the aging population is substantial.6

Smart wheelchair research is not new. A review by Simpson summarizes smart-wheelchair research dating back to 1989 (including efforts at Yale, IBM and Stanford), and includes comparisons of over 50 projects.4 Additionally, Brose and colleagues2 provide a systematic review on various types of robotic assistants for person with disability, including robotic feeding devices and wheelchair arms for those with limited upper limb function. Notable within this body of research is the limited participation by actual users. For example, Brose and colleagues report on case studies for individual users for a number of the more sophisticated robotic systems and “smart” home environments where each system was evaluated by one user. In spite of this history of research in smart wheelchairs, none are in common use. Barriers to acceptance are cost, safety, and ease of use (particularly for older users). In addition, there are few reports of user responses to smart wheelchair features and none that prioritize the features to guide design priorities.

Methods

Design:

Focus group of 5 current wheelchair users from an assisted living facility (ALF). An ALF was chosen as there are potential “smart” wheelchair features (autonomous facility navigation) that were under consideration by the design team.

Sample:

The recreation director at the ALF identified current wheelchair users who were invited to meet with the researchers in a focus group. There were 6 respondents identified, 5 of whom participated. One potential participant was unable to join the focus group because of a health-related issue prior to the focus group.

There were three women and two men; four were white and one was African American. Three of the wheelchair users had both powered and standard wheelchairs and used one wheelchair or the other depending on the setting (inside or outside), distance and terrain (level vs. steep).

The focus group participants were given a flyer explaining the study and asked to let the recreation director know of their interest. Interested participants were given the informed consent document to review. The focus group was held in a private room to decrease the disruption to the focus group participants’ daily routines and to provide confidentiality for participation. The first part of the focus group was a discussion of scenarios with different functions and features. The second part of the focus group was the participant rating of the features on a paper and pencil form for determination of helpfulness and importance. The questionnaire and scenarios were investigator-developed.

Results

For the scenarios, the participants were read the scenarios and then asked “Is this an acceptable function?” The researchers kept field notes of the discussion.

Scenario #1 —

“You have fallen out of bed and cannot get back into bed or get up. The wheelchair responds to your call for help by moving toward you, pointing the camera at you and calling the nurse’s station where the nurse can then direct the camera and speak to you to find out how you are and what you need.

Is this an acceptable function? ”

Four respondents answered yes; one said “not sure” because of questions about whether the wheelchair would be able to get to the respondent through the living environment.

Scenario #2 —

“You have fallen out of bed and cannot get back into bed or get up. You are unconscious. The wheelchair identifies that you have fallen and are not responding to its call. The wheelchair calls for help by moving toward you, pointing the camera at you and calling the nurse’s station where the nurse can then direct the camera and see what’s going on and send help.

Is this an acceptable function?”

All agreed that this was an important function.

Scenario #3 —

“The wheelchair knows your usual daily activities and if you are not up and out of bed by a certain time one morning, the wheelchair will move toward you and ask you if you are ok. If you do not respond, the wheelchair will call the nurse’s station for help.

Is this an acceptable function?”

Two said this would be acceptable; one said that this would be intrusive and two respondents were unsure. Discussion focused on being able to vary one’s routine without the wheelchair being an interruption.

Scenario #4 —

“The wheelchair knows the usual tone and pitch of your voice. The wheelchair sensor determines that your voice indicates you may be having some anxiety or depressive symptoms. The wheelchair will send this information to the central monitoring station where a nurse can follow up.

Is this an acceptable function?”

All agreed that this would be acceptable.

Scenario #5 —

“The wheelchair knows the patterns of your breathing. One day it senses that you are coughing more and are more short of breath than usual. The wheelchair will send this information to the central monitoring station where a nurse can follow up.

Is this an acceptable function? “

All agreed that this would be acceptable with the caveat this would require nurses to be available to review and respond to the reports.

For the second part of the study, the participants were given a paper and pencil survey to complete regarding the helpfulness and importance of the various features of the wheelchair. One user declined to participate in this part of the study.

Discussion

From the scenarios, acceptability was highest for safety issues like falls and integrated telehealth applications for identification of respiratory or mood changes. The lowest acceptability was associated with the wheelchair performing a surveillance activity based on usual patterns of activity, although this was acceptable to several of the participants. The same pattern emerged from the survey--the features ranked as most helpful and most important were primarily related to safety and integrated telehealth applications: fall identification, coughing, illness onset, and communication with the nurse’s station. The lowest priority items were communication with the family and reminders to go to the bathroom. Anecdotally, participants reported that they had other ways to communicate with family and did not think this needed to be a high priority design feature.

There were relatively few concerns about the intrusiveness of the “smart” wheelchair features. Anecdotal comments were consistent with findings reported by Rogers and Fisk7 where concerns about privacy were outweighed by the benefits of the technology. Practical concerns, such as the ability of the “smart” wheelchair to navigate in the apartment or whether there were enough nurses available to receive the monitoring data, were of more concern than loss of privacy. While privacy concerns need to remain on the research agenda, further research may require actual use to determine how the balance between privacy loss and benefits from technology use plays out in the home setting.

While more research is needed into the preferences of older wheelchair users regarding the possible features, results from this pilot study suggest that “smart” wheelchair developers may want to focus development on safety-related and integrated telehealth features. Future research with persons living in independent housing instead of assisted living would be helpful as most older adult wheelchair users live in independent housing. As well, focus groups and interviews with family members (either residing with or apart from the older persons) and health care providers would be informative in prioritization of feature development. Supporting and promoting the independence of wheelchair users, regardless of age, has been a design consideration for many years. The integration of robotics and advanced telehealth sensors can certainly promote continued independence but this but needs to be balanced with user input because of concerns about intrusiveness and the more passive functions of the advanced technologies.

Table 1.

Smart Wheelchair user rating of possible features (N = 4)

Feature	How helpful would this be? 1 = the most helpful feature through 5 = not at all helpful	How important is this to you? 1 = very important, must be included to 5 = not at all important
Integration of telehealth features like a blood pressure cuff, scale for weight	2.25	3.5
A camera for communication with a nurse’s station	1.5	3
A camera for communication with other people like family	4.7	4.7
A reminder system for moving around to avoid skin breakdown from sitting too long	3	3.25
The wheelchair automatically shifts your seat to change the pressure points	1.75	2
Reminders to take your medicines at specific times	3.5	3.5
A medication system that has a drawer that opens when it is time to take medications and tracks whether you have taken it	3.5	3.5
For people who need it, a reminder to go to the bathroom at certain times	3.75	3.75
For people who need it, a reminder to do breathing exercises at certain times and tracks whether they are done or not	3.5	4
Reminders to use the telehealth system to answer questions about how you are feeling	3	3
The wheelchair is your companion or assistant: senses distress and seeks help for you	2.25	2.25
The wheelchair is your companion or assistant: recognizes changes in your daily life patterns and asks how you are doing	3	3
The wheelchair is your companion or assistant: delivers meals or is a courier when you need something and are not using the wheelchair	1.75	2
The wheelchair is your companion or assistant: senses falls and calls for help if you fall	1	1
The wheelchair is your companion or assistant: the camera can be used as a remote monitoring system to check on you	3	3
Senses when you are coughing more or short of breath	1.25	1.25
Senses a change in your voice and tone that may indicate you are sick or something is wrong	1.25	1.25
The wheelchair works with a robot to help you move from the bed to the chair or from the wheelchair to the toilet	2	2
The wheelchair can weigh you and record the information as part of your health monitoring	2	1.5