Designing Assistive Technology For Less-Resourced Environments: An Online Method to Gauge Accessibility Barriers and Collect Design Advice

Jon Pearlman, MS, Alexandra Jefferds, Ila Nagai, BS, HS Chhabra, MD, and Rory Cooper, PhD


An axiom for designing appropriate assistive technology is that users and other stakeholders be involved in the design and development process. When designing devices for developing countries, the designers may be unfamiliar with the remote environment, and stakeholder involvement can be complicated by the stakeholders’ remote locations. To address these difficulties in the design of an electric power wheelchair (EPW) for India, we gathered over 500 photos from wheelchair users who photographed accessibility barriers around their homes in India. EPW experts (users, designers, clinicians, etc.) subsequently reviewed a random subset of the photos (through an online survey), and identified which accessibility barriers are portrayed in the photos, the severity of those barriers, and also provided design ideas for the EPW to allow it to manage those barriers.


participatory action design; developing country; India; electric powered wheelchair


It is estimated that nearly 8 million people need wheelchairs in India [1]. The most common design of manual wheelchairs currently provided is based on older designs not originally meant to be outside an institutational environment [2-4]. While there are efforts to make manual wheelchairs more appropriate for low-income countries like India, little effort has been made to develop an electric powered wheelchair (EPW). In fact, western-style EPWs are beginning to be imported into these countries, which are not likely to meet the needs of the users in their environment. We believe that to avoid the pitfalls of transferring inappropriate technology, it is important to take a pro-active approach to designing devices to be transferred, rather than redesigning a device after it frequently fails. There are different ways to accomplish this. Whirlwind Wheelchairs International (San Francisco, CA, USA), uses open-source design where product designers collaborate with wheelchair builders worldwide to improve and adapt their manual wheelchair design for developing countries. Motivation Charitible Trust (Bristol, UK) uses product designers at its headquarters and in the field to help design their wheelchairs for low-income countries. Our laboratories (the Human Engineering Research Laboratories) use an interative process to improve the design of manual wheelchairs for a large-scale manufacturer in India [2].

A critical element of product design is to ensure the design meets the needs of the users in their environment. This task is difficult when the environment in which the device will be used is not well characterized or known by the designers, which is common when it is foreign. One way to gain an understanding is to interview people from the region, and/or visit the region for extended periods of time [5, 6]. While this effective, it may be difficult to find individuals with thorough knowledge of the region, and may be costly to spend dedicated time in the area; these difficulties are exacerbated when the region is diverse, such as the wide diversity found in the Indian Subcontinent.

As part of a larger project with Indian researchers and AT manufacturers, our laboratories are designing an EPW to meet the growing demand for AT in India. To inform us about the environment that wheelchair users live in and gather design ideas, we have combined a camera study, where wheelchiar users in India are asked to photograph their home and community environments, with an online survey, where a wide range of experts on wheeled-mobility provide feedback on the types of power wheelchair features that should be included in the device, and/or the types of modifications that should be done to the environment.


Phase I-Camera Distribution:

In Spring, 2005, the Indian Spinal Injuries Center (ISIC), in New Delhi, India, recruited 50 wheelchair users to participate in an IRB approved camera study. After informed consent was provided, demographic data were recorded, and subjects were given a disposable camera (28 exposures) with self-addressed envelopes and paid shipping; and a form with instructions on how to use the camera, what to photograph, and lines for the subject to describe each of the photos taken. The subjects were instructed to take pictures in and around their house, work, and neighborhoods where they encountered accessibility barriers. They were also asked to have another person take a picture of them in their wheelchair while performing tasks that were difficult for them. For each photo, they were asked to write down a description of what they were photographing.

After the cameras were returned to the ISIC, they were transferred to the Human Engineering Research Laboratories (HERL) under an exempt IRB approved by the University of Pittsburgh, de-identified and screened. Two wheelchair users were asked independently to screen de-identified photos which did not have obvious accessibility issues displayed (or where there were no accompanying text descriptions). Photos that both screeners agreed did not have accessibility issues were discarded and not included in the dataset.

Phase II- Expert Analysis:

Analysis of the photographic data is currently underway using an online survey approved by the University of Pittsburgh IRB. Subjects are recruited who are knowledgeable about wheelchair use and design, and accessibility issues: wheelchair users and their family members, rehabilitation engineers, service providers (e.g., physical and occupational therapists), and architects who have experience in design and/or modification of environments to make them accessible.


Follow D-JobFigure 1. d (Click for larger view)

Subjects register by filling in a short questionnaire which collection information on their demographic, vocation/occupation, and wheelchair-related expertise. Registration concludes by the subject choosing a username and password to access the site. Once access has been granted, the online system generates a random sequence of 50 images drawn from the entire dataset (~500 images), and presents them one-by-one to the users. Each image is presented with an interactive survey (Figure 1) with questions based on the American Disabilities Act Accessibility Checklist [7], covering issues of steps, rough terrain, doorway widths, ramps, etc. The subjects are asked to choose the accessibility issue portrayed in the photo (13 possibilities), and the severity of the accessibility issue (1- 10 in order of least-to-most severe). The survey includes comment boxes for the subject to describe the power wheelchair design features that would allow the device to accommodate the accessibility barriers portrayed in the photo and/or the changes to the environment that should be made to allow for accessibility.

Statistical Analysis:

Subject demographics from Phases I and II were analyzed with descriptive statistics, as well as the vocational/occupational background of Phase II subjects. Similarly, we analyzed the survey results with descriptive statistics which highlight both the severity of the accessibility barriers, and the frequency with which they were identified in the photos. Reliability of the questionnaire was measured using Interclass Correlation Coefficient (ICC) from repeated photos which were shown to the subject.


Phase I:

Figure 2 shows an example of the image taken by the subjects in India.  It shows an older woman in a depot style wheelchair at her residence.  Someone is standing behind her aiding her in some way.  It is notable that the quality of the photo is not very good.  Also, the accessibility issue the subject is trying to convey is not very clear.  These are common features of the photos throughout the dataset.Figure 2. (Click for larger view)

We have received 30 cameras with a total of 650 photos from the ISIC. After screening was completed, approximately 500 were found to have viable data that displayed accessibility data (e.g., Figure 2). Individuals who returned the cameras included 20 males and 10 females, were 38 (+/-21) years of age, and lived in rural (n=13) and urban (n=17) environments.

Phase II: While this research is ongoing, 50 individuals have been recruited to participate in the online study. The registration collected information on several aspects of the subject’s disabilities, vocation, awareness of EPW design, and awareness of less-resourced environments (Table 1).

Subjects on average reviewed 37% of the 50 photos presented with 18 who completed the entire series, and 3 who reviewed additional photos. A total of 15 subjects viewed 67 repeated images, which resulted in an ICC of .91.

The percentage of overall responses (1100) were distributed across the 13 questions with over 25% related to surface stability (rough terrain, etc.) (Figure 3, light bars). Similarly, subjects rated the severity of the obstacles different, with ‘steps’ being the most severe (Figure 3, dark bars).

Selected comments from subjects are shown (Table 2) to demonstrate some of the design ideas.

Table 1.
Parameter Value Notes
Age 42 (13) Mean (SD)
Gender 25/16 Female/Male
Disability 30 Subjects who either have a disability or who have a family member with a disability
Countries 7/4 Number of different countries subjects were from/ # of total subjects from developing countries
Familiarity With EPWs 3.3 (0.9);
Mean (SD) based on a range indicating no knowledge (1) to expert knowledge (5) of current EPWs; Range of responses
Awareness of Developing Countries 28 # of subjects aware of issues in developing countries
WC Users 30/2/12 # of subjects reporting EPW/Scooter/MWC use (some use multiple devices)
Modifications 21 # of subjects reporting home modifications


Table 2.

large drive wheels and casters, shock absorbers

small outside dimensions, tight turn circle, low knee height

curb climbing capability, dirt/grass capability

close front approach, ability to tuck footrests under bed to move seat as close as possible (an example of how obstacles of varying heights require you to consider 3-dimensional space occupied by wheelchair not merely overall length and width

if client can tolerate scooting on butt on bathroom floor, close front approach to stair might allow transer to floor

Removable seat, or recline to lower the overall height of the wheelchair. Accessible tie-down points.

able to climb at least 4 inches; able to travel over dirt and rocky surfaces independently

swing away footrests or a small foot platform

Deep tread wide rear wheels, wide, eight or 9 inch casters. Suspension system, tilt system, good battery range.

spring release or removable joy stick


Figure 3 is a bar graph showing the frequency and severity rating of the responses to the questionnaire.  For the most part, the frequency that subjects cited certain accessibility barriers was not correlated to the severity that they were rated.  In order of most to least frequent, the subjects cited Surface Stability, Doorway Width, Toilet Height, Objects Approachable, Curbs/thresholds, room under tables/counters, etc.  In the order of most to least sever, the subjects selected, Steps, Bathroom Turning Space, Bathroom Entrance, and ramp, etc.).  These give an indication of how a device should be designed to accommodate the environment.  Figure 3. (Click for larger view)

Participatory design methods [8] are relatively unique in the design of AT, even though they offer tools to help ensure that the devices are appropriate for the end user. We have used these methods to address the widespread problem of inappropriate AT in developing countries, where the devices rarely meet the users’ needs [3, 9, 10]. To address this issue in the development of an EPW for India, we asked wheelchair users in India to photograph accessibility barriers, and had 66 experts from a wide range of countries provide feedback about frequency and severity of those accessibility barriers.

While severity of accessibility barriers in different countries may be similar (e.g., steps are severe in all environments) the frequency of each barrier, we assume, differs significantly. For instance, we assume surface stability would not be the most frequent barrier in and around the homes of wheelchair users in urban United States or European countries. Along with the design ideas provided by the subjects participating in the online questionnaire, the frequency of the accessibility barriers can be important information to consider when designing AT for developing countries—the most widely appropriate device would be able to accommodate the most frequent or prevalent accessibility a barriers.

We received cameras from nearly an even split of males and females, and individuals living in urban and rural environments, which gives a broad but possibly biased view of the barriers in India: the majority of wheelchair users are male, and the vast majority of the population lives in rural environments. We plan to use a multi-site approach for distributing cameras from both rural and urban locations to help ensure the photo database is representative of the overall population. We also will more strongly encourage wheelchair users to write down the accessibility barriers they are photographing, which occurred for less than ½ of the photos we received.

While we had several willing subjects (> 50, with no reimbursement for their time), most did not complete the online questionnaire. Based on comments made by subjects, they did not review all 50 photos because (1) the barriers were hard to identify, (2) 50 photos were too many, and/or (3) the photo quality was poor. We will work to minimize these problems in the future by screening poor photos, encouraging wheelchair users to write what the photo is meant to portray, and reimbursing subjects for their time to complete the questionnaire.

We have plans to expand this study in several ways. To increase the volume and quality of the phase I data, we are looking for collaborative partners in regions world-wide to perform similar camera studies so we can build our database. To increase the versatility of the Phase II methods, we hope to have subjects more directly guide the design—by choosing from a list of potential design features. We also hope to use the database to perform online training of seating and mobility providers.


This work was funded by an NSF IGERT grant (#DGE0333420) and an NSF REU grant (#EEC 0552351). This work would not have been possible without the coordination of clinicians at the Indian Spinal Injuries Center, in New Delhi, India.

  1. India, G.o., Disabled People in India: National Sample Survey (NSS) 58th Round. 2002, Goverment of India: New Delhi. p. 593.
  2. Pearlman, J.L., R.A. Cooper, M. Krizack, A. Lindsley, Y. Wu, K. Reisinger, W. Armstrong, H. Casanova, H.S. Chhabra, and J. Noon, Technical and Clinical Needs for Successful Transfer and Uptake of Lower-Limb Prosthetics and Wheelchairs in Low Income Countries. IEEE-EMBS Magazine, 2006 (In Press).
  3. Mukherjee, G. and A. Samanta, Wheelchair charity: a useless benevolence in community-based rehabilitation. Disabil Rehabil, 2005. 27(10): p. 591-6.
  4. ALIMCO, Articial Limbs Manufacturing Corporation of India 2003 Annual Report. 2003, Indian Government: Kanpur, India.
  5. Mulholland, S.J., T.L. Packer, S.J. Laschinger, J.T. Lysack, U.P. Wyss, and S. Balaram, Evaluating a new mobility device: feedback from women with disabilities in India. Disability & Rehabilitation, 2000. 22(3): p. 111-22.
  6. Mulholland, S.J., T.L. Packer, S.J. Laschinger, S.J. Olney and V. Panchal, The mobility needs of women with physical disabilities in India: a functional perspective. Disabil Rehabil, 1998. 20(5): p. 168-78.
  7. NIDRR, The Americans With Disability Act Checklist for Readily Achievable Barrier Removal (for Existing Facilities), D.o. Eduction, Editor. 1995, Barrier Free Environments, Inc. Adaptive Environments Center, Inc. p. 1-15.
  8. Schuler, D. and A. Namioka, Participatory Design: Principles and Practices. 1993: Lawrence Erlbaum Assoc Inc.
  9. Singhal, D. and S. Nundy, No Mean Feet., 2004. 328: p. 789-.
  10. Saha, R., A.K. Dey, M. Hatoj and S. Poddar, Study of Wheelcahir Operations in Rural Areas Covered Under the District Rehabilitaiton Centre (DRC) Scheme. Indian Journal of Disability and Rehabilitation, 1990: p. 74-87.


  • Source Ordered
  • No Tables
  • Very Compatible


Disney produced a television show in the mid 1990s called Gargoyles. It's a great show and I'm a big fan. A few years ago Disney started to release the show on DVD. The last release was of season 2, volume 1. That was two years ago. Volume 2 has not been released. Why? Poor sales. So if you should find yourself wanting to support my work, instead I ask you pick up a copy of season 2, volume 1. It's a great show and you might find yourself enjoying it.