Revisiting Clear Floor Area Requirements for Wheeled Mobility Device Users in Public Transportation

Abstract

Current accessibility standards in the U.S. prescribe minimum dimensions for ‘clear floor area’ to accommodate wheeled mobility device (WhMD) users on transportation vehicles. Prior research on the anthropometry of WhMD users (n = 500) indicates that these dimensions are too small to accommodate the size of many occupied wheeled mobility devices, especially power chairs and scooters.

This paper describes a development project designed to update the evidence for these technical criteria and communicate them to the vehicle designers and accessibility standards developers in a manner that would facilitate making good decisions. An interactive web-based design tool was developed for determining the dimensions of clear floor area to achieve a user-specified level of physical accommodation based on occupied device length and width measurements taken on 500 WhMD users. The web-based design tool is now available to practitioners who seek to accommodate a wider range of WhMD users than the minimum standards required by regulations.

The design tool is also intended as a visual evidence base for regulatory activity and universal design practice with higher ambitions. The advent of driverless automated vehicles will increase the importance of accessibility and usability to accommodate the diversity of riders with disabilities. Clear floor space to enable independent ingress, interior circulation and egress among WhMD users will be a foremost concern. The transportation industry, standards developers, disability advocates, mobility device manufacturers and prescribers need to understand the limitations of current accessibility standards and work to address these limitations through updated vehicle design standards and policies.

INTRODUCTION

Increases in life span, improved medical care, and increasing availability of assistive technologies that promote independence have resulted in a growing need for accessible public transportation that provides efficient and safe services for those with disabilities, including users of wheeled mobility devices (WhMDs). The US population of non-institutionalized (i.e., not in nursing homes, hospitals) WhMD users is currently at 3.4 million (1) and continues to grow at an annual rate of 5% (2). However, over 40% of WhMD users living in areas served by public transit experience significant environmental and design barriers when using public transit services (3). WhMD users present a unique set of challenges for physical accessibility in public transportation because of concerns over large spatial requirements for floor area and maneuvering (4, 5).

Currently, accessibility standards and codes are used throughout the U.S. and in other countries to implement laws that mandate accessibility to transportation (e.g. buses, vans, trains, etc.) and buildings (e.g. public restrooms, bus stops, transit facilities, etc.). The Americans with Disabilities Act Accessibility Guidelines (ADAAG) for Transportation Vehicles (6) is the key document used for design of accessible facilities and transportation systems. A primary feature of the ADAAG is the specification of a ‘clear floor area’ for wheelchairs, a space 760 mm (30 in.) wide by 1220 mm (48 in.) in length. The ‘clear floor area’ is used to determine the size of the area planned for wheelchair securement, the width of doorways and minimum interior clearances for maneuvering a wheeled mobility device, and the space needed to accommodate people in wheeled mobility devices in bus shelters, bus stop pads and in terminals - hence, a critical component for ensuring physical accessibility.

The technical criteria in these standards are based on body sizes and functional abilities (i.e., anthropometry) of adults and sometimes children with disabilities. To understand the spatial implications of contemporary wheeled mobility technology and user populations, the second author (CD) collaborated with researchers at the IDeA Center at the University at Buffalo to develop an extensive anthropometry database of wheeled mobility devices in the U.S. with the intention of improving accessibility to buildings, facilities, and public transportation vehicles (7, 8). A key finding from an early analysis on the anthropometry of 369 WhMD users indicated that the minimum clear floor area dimensions for transit buses are too small to accommodate the size of many occupied wheeled mobility devices, especially power chairs and scooters (9). Key reasons for this discommodation are worth revisiting. Foremost was the realization that research conducted in the 1970’s was the basis for the current criteria for accommodating WhMD users (10). Since that time, the sizes and characteristics of WMDs and their users have changed considerably. By the late 1990’s, anecdotal evidence was emerging that the standards were no longer adequate.

Today, the range of WhMDs available is much more diverse. The size, weight and manoeuvring characteristics of wheeled mobility devices vary considerably between device type (e.g., manual wheelchairs, powered wheelchairs and electric scooters), make, and model (7–9,11). Both manual and powered wheelchairs offer some customizability to fit user needs in postural support, comfort, and ease of use (e.g., leg rests and footrests, tilt and recline) which also impacts space requirements (reference). The growing availability and use of scooters and power chairs is expected to further change the mobility device user demographics in the coming years (2). Power chairs have allowed more severely disabled people to become mobile in the community. The choice of drive configuration (i.e., front, mid, and rear wheel-drive) and powered seating options influence manoeuvrability of powered wheelchairs (12). People are larger due to the obesity epidemic resulting in a growth in bariatric wheelchairs.

Increasing use of scooters among the growing older adult population has put greater emphasis on scooter access in public transit systems. Scooters typically have a larger turning circle and are less manoeuvrable than wheelchairs because their configuration and chassis length need to accommodate the drive controls (e.g., tiller) and foot placement (12–14). These devices tend to be designed and prescribed for outdoor use. They are intended to assist individuals with limited ambulation and good trunk control and sufficient upper extremity function to control the steering tiller, but who also lack the upper extremity stamina or range of motion necessary to use a manual wheelchair. These problems are not unique to the US. Research studies from the UK suggest similar concerns regarding increased wheeled mobility sizes and challenges to accommodation on public transit vehicles (7, 15, 16).

Over time, the emphasis of our research effort evolved to creating data-driven graphical design tools to provide an evidence base to evaluate existing accessibility standards and to support designers and standards developers who seek to accommodate a wider range of WhMD users than the minimum standards required by regulations (17, 18). Facilitating this outcome was the production of statistical analyses and visualizations with re-sults in a graphical format (e.g. percentage of WMD users that could fit, turn, reach, see, in specific conditions). In particular, multivariate accommodation models were very effective for communication and standards development activities; they provided a sophisticated way to visualize the impact of revisions to accessibility standards as “what if” analyses (17, 18). This shift was informed by lessons learnt from our engagement with the design and accessibility standards community (18) that:

• Research results are best presented to decision makers in a form that is familiar to them, in this case, two dimensional illustrations that resemble or can be overlaid with illustrations used in the existing standard, and,

• Graphic presentations that illustrate the impact of alternative decisions, like interactive accommodation models, can greatly facilitate adoption of findings.

Study Objectives

The objectives of this study were to:

1. Develop a data-driven algorithm for determining the dimensions of clear floor area to achieve a user-specified level of physical accommodation based on occupied device length and width measurements taken on 500 WhMD users,

2. Implement the algorithm as a web-based accommodation model with a graphical and interactive user interface to support standards development, regulatory action, and design,

3. Evaluate existing clear floor area requirements prescribed in the ADAAG using the developed accommodation model and where necessary generate recommendations for improving the minimum clear floor area requirements.

   We also discuss the implications of our finding.

METHODS

The analysis described in this study used anthropometry measurements previously taken on 500 adults in the US who relied solely on WhMDs for mobility (8). The WhMD users were recruited through many sources, including a local independent living center, a United Cerebral Palsy Association, a geriatric day care center, and local hospitals, including Veterans Affairs Medical Centers in Buffalo and Pittsburgh. In addition, advertisements were posted in local newspapers and flyers placed in local organizations and stores. The university’s institutional review board approved the study and all participants provided written informed consent prior to participation.

The research team made a deliberate attempt to recruit a diverse group of users, rather than just individuals who possessed a specific set of physical capabilities so that the results obtained could be extended to the broader wheeled mobility device user population in the US. The measurement protocol included the collection of demographic information, mobility device specifications, structural anthropometric measurements and functional anthropometric information from each participant (8).

A three dimensional digitizing technique was developed to collect detailed data on static anthropometric variables and put the data into a form that could be used for a variety of different purposes (19, 20). A set of over 130 standardized three-dimensional data points for each person and device were collected that could describe the three-dimensional volume a person and device occupied, while individuals held a comfortable or typical seated posture. Three-dimensional coordinates of the landmarks were then used to derive estimates of widths, heights and depths of mobility device characteristics and body dimensions relevant to built environment design. The current study used three computed dimensions:

1. Occupied length: computed as the horizontal distance from the extreme rear-most and forward-most point of the combined occupant or mobility device.

2. Occupied width: was computed as the horizontal distance between the extreme lateral right-most and left-most points on the body or mobility device with the participant seated in a comfortable, relaxed posture.

3. Occupied floor area: defined as the the estimated rectangular floor space required for each occupant-mobility device in the study sample and computed as a product of the occupied length and occupied width.

WEB DATABASE AND USER INTERFACE

This development project advances our anthropometry research by implementing a new web-based framework to support multivariate analysis of specific aspects of WhMD user interaction with the built environment (e.g., clearance, approach, reach capability, line of sight, and turning space requirements). This framework consists of an interconnected, hierarchical set of modules (using open source scripting languages PHP and JavaScript) to query the underlying anthropometry of wheeled mobility database implemented in MySQL and to perform for multivariate anthropometry analyses. A visual interface was developed using D3.js, which is a JavaScript library for producing dynamic, interactive data visualizations in web browsers by combining SVG for graphics, HTML5 for content, CSS (Cascading Style Sheets) for aesthetics, PHP for data retrieval, and Javascript for data manipulation.

Displaying Clear Floor Area

The visual interface allows users to retrieve and display information on occupied width and occupied length as individual datapoints on a two-dimensional scatter-plot (Figure 1). A scatter-plot is very useful for depicting the association between two variables, i.e., occupied length and width in our case. Data can be displayed for all 500 WhMD users in the study sample or stratified by a defined sub-group (e.g. type of WhMD, sex, age range, etc.).

FIGURE 1.

Scatter-plot of occupied length vs. occupied width, overlaid with the ADAAG Sec. 305.5 requirement for minimum ‘clear floor area’ of size 760 mm × 1220 mm (30 in. × 48 in.).

To provide context to designers and standards developers, a plan view of a WhMD user and ADAAG minimum requirement for clear floor area (reference) is superimposed on the display. Datapoints enclosed by the ADAAG required minimum clear floor area (Quadrant 1) represent individuals accommodated, while datapoints outside represent individuals excluded either due to inadequate clear floor length (Quadrant 2), width (Quadrant 3), or both length and width (Quadrant 4).

Accommodation Model for Clear Floor Area

An accommodation model was developed for determining the minimum dimensions of clear floor area to achieve a user-specified level of physical accommodation based on occupied device length and width measurements taken on 500 WhMD users. In the interactive web-based display two modes for conducting “what if” analyses are implemented. First, end-users can drag the boundary of the clear floor area (dotted lines in Figure 1) to increase/decrease the clear floor area with the interface display updating in near-realtime to communicate the proportion of the sample accommodated (i.e., number of datapoints in Quadrant 1 divided by the total number of datpoints for that group of users). Alternately, the end-user can specify on a slider-bar the percentile of users to the accommodated and the boundary of the clear floor area on the 2-D display updates in near-realtime to depict the corresponding minimum required clear floor area length and width.

The latter percentile analyses occupied floor area provides a more useful starting point for determining suitable dimensions for ‘clear floor area’. In contrast to our previous report (9), this study presents a more refined method for computing the minimum required clear floor area length and width to accommodate a desired percentile of WhMD users. The need for this stems from the fact that the Pearson’s correlation coefficients between occupied width and occupied length were weak, i.e., 0.21 for manual chairs, 0.30 for power chairs, and 0.50 for scooters, suggesting that neither occupied width or occupied length can necessarily be used to reliably predict the other. It also implies that the specific combination of clear floor area length and width to accommodate a desired percentile is not trivial.

Procedure for Calculating Minimum Clear Floor Space Requirements by Percentile

Summarized below is the procedure developed for calculating by WhMD type the minimum clear floor area length and width to accommodate a user-specified percentile accommodated and is also depicted in Figure 2:

FIGURE 2.

Quadratic fit for least areas stratified by mobility device type, overlaid with the ADAAG Sec. 305.5 requirement for minimum ‘clear floor area’ of size 760 mm × 1220 mm.

1. For a given percentile of accomodation, say 5%, the WhMD users in the study sample whose corresponding occupied length and occupied width accommodate 5% ± 2% of users in the sample are identified. We opted for a tolerance range of ± 2% due to the relatively small sample sizes, particulary for scooter users. This step produces multiple possible lengths and widths for a given value of percentile accommodation.

2. Next we determine the specific combination of length and width that would produce the smallest clear floor area from these identified points. This is done by fitting or approximating a quadratic curve to the identifed data points (i.e., a yellow curve at 5% in figure 2). Any point on this curve would accommodate 5% of users; some combinations with large occupied widths, some with large occupied lengths, while others in-between. A quadratic curve was chosen due to low mean squared error values compared to other polynomial and linear curves, i.e., a low-order non-linear relationship. The equation for a quadratic curve takes the following general form,

   $$f(x)=y=a{x}^2+bx+c$$ (1)
3. The clear floor area obtained by any point on the curve can be represented as,

   $$\mathit{\text{Area}}\text{(}A\text{)}\text{=}xy$$ (2)

   where ‘x’ is the occupied length and ‘y’ is the occupied width.
   Substituing for ‘y’ from Eq. (1),

   $$A=x\left(a{x}^2+bx+c\right)$$ (3)
   The point (x_min, y_min) on this quadratic curve that would yield the smallest clear floor area can be obtained by taking the first derivative, equating to zero, and finding the solution

   $$\frac{d}{dx}(A)=\frac{d}{dx}\left(x\left(a{x}^2+bx+c\right)\right)=0$$ (4)

   $$3a{x}^2+2bx+c=0$$ (5)
4. The solution to this equation is :

   $$x_{\mathit{\text{min}}}=\frac{2b\pm \sqrt{4b^2-12ac}}{6a}$$ (6)
   Substituting for x in Eq. (1), we obtain :

   $$y_{\mathit{\text{min}}}=a{\left(x_{\mathit{\text{min}}}\right)}^2+b\left(x_{\mathit{\text{min}}}\right)+c$$ (7)

5. Steps 1 to 4 are repeated for all percentiles from 5% to 95% in increments of of 1%. Figure 2 shows the corresponding quadratic curves in increments of 5%.

6. Next, a quadratic curve is fit through all the pairs of minimum occupied length and occupied width obtained from each percentile value. Again, a quadratic curve is chosen due to low mean squared error. Each point on this curve provides a clear floor area with a corresponding length and width and this area will be close to minimum for a given accommodation value. This is depicted by the black curve in Figure 2 passing through the different percentile-based combinations of points (x_min, y_min). The curve shows the trajectory by which the occupied length and width should change to increase/decrease accommodation. Clearly, this trajectory is non-linear, i.e., occupied length and width do not change proportionately. Separate trajectories were produced for each type of WhMD.

DATA ANALYSIS

Three types of analyses were performed stratified by the type of WhMD:

1. Descriptive analyses to evaluate the distributional characteristics of occupied length, width, and area. The mean, standard deviation, range, 5^th and 95^th percentile values for occupied length, width, and area were calculated using SPSS Statistics v23.0 (21).

2. Occupied lengths and widths in the sample were compared to the ADAAG minimum requirement for clear floor area to determine the proportion of the sample accommodated and disaccommodated.

3. Using the graphical accommodation model described in the previous section, the minimum required clear floor area lengths and widths for commonly used percentile accommodated were computed and summarized in graphical and tabular format for use by accessibility designers and standards developers.

RESULTS

Sample Demographics

The study sample consisted of 264 (52.8%) men and 236 (47.2%) women. Among men, 147 (55.7%) were manual chair users, 100 (37.9%) power chair users, and 17 (6.4%) scooter users. Among women, 130 (55.1%) used a manual chair, 89 (33.7%) used a power chair and 17 (7.2%) used a scooter. The mean (standard deviation) age of the sample was 56.4 (18.8) year with a range of 18 to 100 years.

Regarding self-reported medical conditions that led to dependence on mobility devices, CNS disorders (e.g. multiple sclerosis, cerebral palsy, etc.) were the most frequently occurring (34%), followed by spinal cord injuries (26%), orthopedic injuries/deformities (13%) and cerebral vascular diseases such as stroke (10%). Instances of medical conditions such as amputations (3%), traumatic brain injuries (2%), or respiratory and cardiovascular diseases (2%), were less frequent, with the remaining cases listed under ‘Other’ (10%).

Univariate analysis of occupied width, occupied length and occupied floor area

The descriptive statistics for occupied width, occupied length, and occupied floor area stratified by mobility device type are summarized in Table 1.

TABLE 1.

Summary Statistics For Occupied Width (mm), Occupied Length (mm) And Occupied Area (m²) Statrified By Type of Wheeled Mobility Device.

Manual chairs (n = 277)
Dimension	Mean	SD	Min	Max	Percentile
					5th	50th	95th
Occupied Width (mm)	685.0	61.4	508.0	992.0	595.7	677.3	786.4
Occupied Length (mm)	1150.5	129.8	743.0	1625.0	934.7	1154.8	1362.0
Occupied Area (m2)	0.79	0.125	0.44	1.25	0.60	0.78	1.00
Power chairs (n = 189)
Dimension	Mean	SD	Min	Max	Percentile
					5th	50th	95th
Occupied Width (mm)	706.7	72.4	574.0	1008.0	606.9	695.3	827.5
Occupied Length (mm)	1196.3	133.7	831.0	1709.0	977.0	1183.4	1414.5
Occupied Area (m2)	0.85	0.145	0.50	1.27	0.63	0.84	1.11
Scooters (n = 34)
Dimension	Mean	SD	Min	Max	Percentile
					5th	50th	95th
Occupied Width (mm)	650.0	89.6	488.0	857.0	526.7	617.3	829.3
Occupied Length (mm)	1208.5	110.0	1025.0	1439.0	1039.3	1202.8	1433.1
Occupied Area (m2)	0.79	0.158	0.50	1.17	0.55	0.77	1.12

Comparisons with the ADAAG Minimum Clear Floor Space Requirement

Table 1 provides the frequency count and percentages in parentheses (percentages are computed based on the corresponding mobility device category) of the number of cases in each of the four quadrants depicted in Figure 1. Only 59.4% of combined mobility device users in the sample were accommodated in the ADAAG minimum clear floor space requirement (Quadrant 1). The analysis of occupied length and occupied width of mobility device users in the study sample with the ADAAG prescribed minimum clear floor length (1220 mm) and minimum clear floor width (760 mm), shows that a larger percentage of mobility device users in the study sample, exceeded in occupied length (Quadrant 2) as compared to occupied width (Quadrant 3) or both, occupied length and width (Quadrant 4)

Percentile-based Accommodation Model for Clear Floor Area Length and Width

Figure 3 show the quadratic fit for the data points at which least area for a given percentile is observed in the sample. The coefficient of determination (r²) or the fit of the quadratic curve is determined to be 0.86 for all mobility devices combined, 0.89 for manual chairs and 0.80 for power chairs. Scooters had a very low r² value due to insufficient data points that result from small number of scotter users in the sample. Table 3 gives the same values stratified by each device type (manual chair and power chair).

FIGURE 3.

Accommodation model for Clear Floor Area by Type of Wheeled Mobility Device. Close up on the right.

TABLE 3.

Percentile Accommodated and the Corresponding Estimated Minimum Clear Floor Area Length and Width Stratified by Mobility Device Type

Dimension	Percentile Accommodated for Manual Wheelchair Users (n = 277)
	5%	15%	25%	35%	45%	55%	65%	75%	85%	95%
Length(mm)	1008	1111	1163	1195	1230	1255	1295	1319	1352	1449
Width (mm)	658	660	670	679	692	702	721	734	754	828
Area (m2)	0.66	0.73	0.78	0.81	0.85	0.88	0.93	0.97	1.02	1.20
Dimension	Percentile Accommodated for Power Chair (n = 189)
	5%	15%	25%	35%	45%	55%	65%	75%	85%	95%
Length (mm)	1076	1168	1204	1244	1267	1320	1338	1355 *	1372	1508 *
Width (mm)	616	652	669	689	702	732	744	755 *	766	870 *
Area (m2)	0.66	0.76	0.80	0.86	0.89	0.97	0.99	1.02	1.05	1.31
Dimension	Percentile Accommodated for Scooters (n = 34)
	5%	15%	25%	35%	45%	55%	65%	75%	85%	95%
Length (mm)	1148	1161	1246	1252	1253	1291	1310	1327 *	1343	1360 *
Width (mm)	549	548	609	617	619	687	729	785 *	817	860 *
Area (m2)	0.63	0.64	0.76	0.77	0.77	0.89	0.95	1.04	1.10	1.17

^*Indicates values that were interpolated or extrapolated from the best-fit quadratic curve due to few observations at those percentiles.

DISCUSSION

Contemporary wheeled mobility technology can support greater independence, better utilization of public transportation and increased social participation of individuals with disabilities into mainstream society (e.g. employment, shopping/doctor visits, recreational travel). However, the full potential of this technology cannot be realized unless device size and shape, and the size of occupants are accommodated by transportation systems to insure access to safe, comfortable and timely travel.

Implications for Accessibility Design Standards and Policies

Standards serve as the foundation for evidence-based design of buildings because architects, product designers, and other design professionals rely on experts in accessibility, safety, and public health to translate knowledge into guidelines for practice. Like product development, standards development is a long term, expensive, and time-consuming activity. It requires the creation of supporting evidence and translation of that evidence into realistic standards.

The findings of this research provide a guide to policy makers and accessibility standards developers for evaluating and revising current transportation accessibility standards. The results suggest that the current standards are not adequate to accommodate a large proportion of contemporary wheeled mobility users. Three design strategies could address this finding in practice. The standards for minimum clear floor area could be increased. This would have limited, if any, implications for new construction in the built environment but could have major implications for design of some vehicle types, especially low floor buses where the space between wheel wells is very limited. The use of a four-passenger longitudinal flip-up seat in place of a three-passenger longitudinal flip-up seat in one wheelchair securement area would provide enough space for larger scooters and wheelchairs, reducing overall seating capacity by only one person. Finally, location of boarding ramps at entries in the middle of low floor vehicles instead of front door access could increase the maneuvering space available by avoiding the need for passage between wheel wells. This last strategy would require using a curb-side fare collection or an alternate method that does not require face to face driver interaction or supervision. It also may have, at present, a limited application to boarding from raised platforms because in most low floor buses, the kneeling feature, which enables a lower ramp slope, is only installed on the front of the bus. These strategies above are not mutually exclusive. Together, they provide policy makers with some options that could lead to a reasonable solution with minimum cost and operational impact. The alternative is to enact operating policies that restrict access on mass transport vehicles to devices that have footprints that fall within specified limits, but such policies based on size could be considered discriminatory and require drivers to make difficult judgments in the field.

Implications for Accessible Design of Transportation Vehicles

Small increases to the minimum securement area to better accommodate contemporary mobility devices on occupied length and width also lead to reduced entry and positioning times and depending on the specific bus design without affecting seating capacity (22). However, it is also important to note that a space increase in existing public transit environments may be prevented by technological constraints.

The results presented here are not only useful for improving regulations and policies but also to identify design goals for future transportation and wheeled mobility technology. For example, the data provide a goal for design of new suspension systems for low floor buses that could increase the available space between wheel-wells, for new securement systems that might allow additional available room in the securement area, or alternative vehicle interior layouts that offer better access and circulation spaces. Also important, in this regard, are innovations in mobility device design that improve device maneuverability making them more adaptable to constrained environments such as on buses, or technologies that increase the comfort and postural support for the occupant while possibly reducing the occupied floor area.

Designing the interior of buses and other transit vehicles for accessibility is a multidimensional design problem. Multiple adjacent design features in addition to the size of the securement area such as its location in relation to the front or rear-doorway, forward vs. rear-facing orientation) and seat layout and orientation (transverse vs. longitudinal) influence an individual WhMD users’ ability to enter, situate in and exit the securement area in an efficient manner. In some instances, it can require simultaneous consideration for additional variables such as shoulder widths and heights, knee and toe clearance widths and heights and functional reach capability to name a few. Future research activities will explore the development of multivariate analyses tools that can factor in additional variables besides width and length of the occupant and device. Additionally, our research group has conducted laboratory-based experiments using a full-scale mock-up of a bus with reconfigurable interiors the relationship between occupied device dimensions and actual performance in dynamic activities such as boarding, interior circulation and disembarking (22, 23).

The aging of the population, rising social expectations of accessible mobility options, and the increase cost of vehicle ownership will lead to increase the utilization and demand for inclusive, reliable transportation systems. The advent of automated vehicles will present opportunities for reducing mobility barriers for individuals with disabilities. But little attention is being given to the design of these vehicles to serve the disability community. The lack of drivers will also introduce new problems. On the one hand, the design of automated vehicles will necessitate a higher standard of usability to insure independent use without driver assistance. On the other hand, such vehicles will have more space due to the removal of obsolete components such as driver stations and fare machines. The new mobility industry and technology developers have yet to embrace the need and value of designing these vehicles to be accessible and inclusive. Standards developers, disability advocates, mobility device manufacturers and prescribers need to understand the limitations of current accessibility standards and work to address these limitations through updated vehicle design standards and policies. As demonstrated in the case example of clear floor area, human factors and ergonomics research can improve our knowledge of how best to accommodate the diverse population of passengers.

Methodological Limitations

Three study limitations are worth emphasizing. First, the proportion of wheelchair users mentioned above might not represent the same proportion expected in adult U.S. population of wheelchair users. For example, there are an estimated 84% manual chair users, 8% power chair users, and about 8% scooter users in the adult U.S. population of wheeled mobility device users (2, 3). Compared to these proportions, the current study sample had a much larger proportion of power wheelchair users and more severe physical limitations compared to the U.S. population. However, power chair users were intentionally oversampled to obtain a better understanding of the functional abilities of this user group, which typically has more severe physical limitations and is hence more sensitive to design restrictions.

Second, measurements in this study were taken with participants seated in a comfortable posture that they could maintain for the duration of the measurement process, typically lasting 15–20 minutes. Participants often chose to rest their arms on the armrest and/or extend the footrests and legrests to support the lower extremities. This data does not take into account the ability of some individuals to move their limbs inboard of their devices, swing legrests and footrests out of the way, or adjust back-packs and bags which may extend beyond the boundaries of their devices. It is also worth noting, however, that traveling on public transportation is an activity that requires remaining relatively stationary for extended periods of time. While these measurement conditions and postures may tend to overestimate the occupied width and occupied length dimensions for very short time durations, they do reflect postures that wheelchair occupants consider comfortable for times that are commensurate with the vast majority of transit trips. Further, the samples were drawn largely from cold weather cities which may introduce some bias toward larger and more durable equipment, but data was collected all year round in an attempt to minimize any bias in recruitment associated with the seasons.

Last, the analyses presented in this paper are based on the assumptions of a rectangular area required for static positioning of a wheeled mobility device. The findings suggest that the current minimum requirements of a 760 mm (30 in.) wide by 1220 mm (48 in.) length of ‘clear floor area’ are clearly inadequate for a sizeable proportion of wheeled mobility device users measured in this study, especially those with the most serious disabilities, namely, power chair users. Wheeled mobility anthropometry dimensions from this study have also been compared with existing accessibility standards and anthropometry research findings from other countries such as Australia, Canada and U.K. (7). These international comparisons also suggest a trend toward an increase in occupied floor area among mobility devices, and the inability of accessibility standards to reflect the space requirements of wheeled mobility device users in the built environment. One approach to increasing the level of accommodation would be to provide a larger clear floor area than the existing standards. The findings described here are intended to provide guidance for modifying standards, particularly in applications where the amount of space available in some vehicles is very limited. A large number of individuals in the study were excluded by the minimum clear floor area requirements due to larger occupied lengths (Quadrant 2 in Figure 1), suggesting that the greatest improvements in accommodation could be first achieved by increasing the length dimension of the ‘clear floor area’. It is worth noting that based on our research findings and other considerations from stakeholders, the ICC/ANSI A117.1 Committee, which publishes a voluntary building standard for accessibility has revised its minimum requirements for clear floor area length in all new construction, increasing it from 1220 mm (48 in.) to 1320 mm (52 in.) (18, 24). Also, space requirements for maneuvering into and out of these spaces inside vehicle were not considered in the current analysis and might suggest the need for additional clearances (22).

CONCLUSION

This report describes a data-driven graphical and interactive accommodation model that can be used by architects and designers to visualize and understand the impact of physical design features on accommodation levels. The accommodation model presented here shows the importance of linking occupied width and length of wheeled mobility devices when establishing suitable dimensions for minimum ‘clear floor area’ in accessibility standards, including those used in the transportation sector. The findings from the analysis of clear floor area need to be the subject of considerable dialogue in the transportation community to determine what level of accommodation is realistic and practical with regard to contemporary vehicle technology. Involving wheeled mobility aid manufacturers, mobility device vendors, rehabilitation engineers and therapists, and representatives from transit agencies and organizations serving individuals with disabilities in this dialogue is also very critical to arriving at acceptable and easily implementable solutions. For now, given prevailing standards provisions, rehabilitation engineers and therapists prescribing wheelchairs need to consider the transportation needs of clients and the occupied device size relative to the space available for wheelchair circulation and securement areas on contemporary public transportation vehicles.

The report also demonstrates the utility of the assembled anthropometry database in helping standards developers and designers improve accessibility, independent mobility, and safety for travelers using wheeled mobility devices on buses, other public transportation vehicles and potentially future autonomous driverless vehicles. The web-based implementation provides the accessible design and standards community world-wide with the capability to query our anthropometry database and perform multivariate analyses without requiring users to have programming or statistical expertise. Usability testing of the website and implementation of other such multivariate accommodation models is ongoing. The accommodation model can be accessed via the website for the University of Michigan’s Inclusive Mobility Research Laboratory at http://dsouzalab.engin.umich.edu/research/anthro/index.php.

TABLE 2.

Percentage Of Wheeled Mobilty Device Users in the Study Sample (n = 500) With Lengths And Widths That Are Accommodated (Quadrant 1) Or Exclusded (Quadrants 2, 3, 4) Compared to the ADAAG Required Minimum Clear Floor Area Of Size 760 mm × 1220 mm.

Mobility Device Type	Quadrant				Total
	Q1 (Included) L<1220 & W<760	Q2 (Exceed length) L>=1220 & W<760	Q3 (Exceed width) L<1220 & W>=760	Q4 (Exceed length & width) L>=1220 & W>=760
Manual	185 (66.8%)	65 (23.5%)	15 (5.4%)	12 (4.3%)	277 (100%)
Power	94 (49.7%)	55 (29.1%)	16 (8.5%)	24 (12.7%)	189 (100%)
Scooter	18 (52.9%)	12 (35.3%)	1 (2.9%)	3 (8.8%)	34 (100%)
Combined	297 (59.4%)	132 (26.4%)	32 (6.4%)	39 (7.8%)	500 (100%)