Accessible Design of Low-Speed Automated Shuttles: A Brief Review of Lessons Learned from Public Transit

Abstract

Low-speed, driverless automated shuttles have the potential to significantly improve community mobility for older adults and people with disabilities who are otherwise unable or ineligible to drive. However, accessibility and inclusive design of these shuttles to accommodate the spectrum of human abilities and impairments is impeded by the lack of accessible design regulations, standards, and information tools specific to automated vehicles. In light of the scarce accessibility research on automated shuttles, a literature review on public transit was performed. A total of 66 documents were identified addressing components of the public transit travel chain involving older adults and people with disabilities. This paper reviewed 11 of the documents pertaining to onboard circulation. Findings highlight the importance of vehicle interior design on accessibility and usability for people with impairments, the inadequacy of existing accessibility standards when designing beyond minimum requirements, and a lack of evidence-based design tools and information to support designing for accessibility. An interactive web-based repository for transportation accessibility research is introduced to inform the accessible design of LSAS, along with directions for future research.

INTRODUCTION

Automated vehicles (AVs) have received considerable attention and investment in recent years with the potential to transform mobility for all. Low-speed automated shuttles (LSAS) are a category of driverless (Level 4 & 5), low-floor AVs designed to carry 4 to 15 people depending on the specific design, operate at slow speeds (10 to 25 mph), and serve short distance trips in shared use settings such as on hospital and university campuses, airports, downtown corridors, and residential communities (Cregger et al., 2018). Nearly a dozen communities in the US and an even larger number internationally have deployed either pilot or fully operational LSAS such as the Olli (Local Motors Inc., Arizona), Arma (Navya Inc., France), and EZ10 (Easymile Inc., France) (Cregger et al., 2018).

Early-use cases for LSAS suggest a momentous opportunity for promoting safe and independent mobility among older adults and people with disabilities who are unable or ineligible to drive (Bradshaw-Martin & Easton, 2014; NCD, 2015). Approximately one-fifth of the US population has some type of physical, visual, auditory, or cognitive disability (US Census Bureau, 2016). Disability prevalence is also known to increase with age (Taylor, 2018). A growing body of research indicates that people with disabilities continue to face transportation barriers leading to decreased social participation and overall well-being (Broome, McKenna, Fleming, & Worrall, 2009; Delbosc & Currie, 2011). A number of these barriers stem from accessibility and usability issues with using public transit vehicles (Audirac, 2008). Realizing the full potential of shared-use LSAS to address current barriers in safe and independent community mobility will require providing designers and policymakers with the information and tools to generate and evaluate designs that consider the needs of diverse range of human abilities and impairments along different phases of the travel chain.

Designing LSAS for universal accessibility is currently challenged by the absence of accessibility regulations, standards, and best practices specific to driverless AVs (NCD, 2015). In addition, design aids and information derived from human factors research needed to support accessible design in this domain are lacking. Recent reports highlight the need for a collaborative effort to research, review, and develop recommendations for updated standards, policies, and regulatory frameworks in order to ensure both sustainable deployment and universal accessibility within an AV transportation system (NCD, 2015; Cregger et al., 2018). This reflects the need to create a centralized repository for accessible transportation research to facilitate communication between researchers, policymakers, and industry on current knowledge, challenges, and needs. As part of a larger on-going research effort to develop accessibility guidelines for LSAS, a literature review was conducted to identify and integrate prior accessibility research on existing ground-based public transportation systems that may be applicable to shared-use LSAS, and to identify knowledge gaps and directions for future accessibility research related to LSAS.

The objectives of this paper were to (1) summarize results from a portion of the literature review that relate to onboard interior circulation, (2) present an interactive, scalable, web-based repository to capture prior research based on the review and future transportation accessibility research. The repository was developed to serve as an information resource that can be continually updated to support the accessible design of LSAS. Implications and directions for future research are discussed.

METHODS

A systematic search on GoogleScholar and the University of Michigan Library Database Search which includes databases for the Transportation Research Board’s Transport Research International Documentation, Scopus, and Web of Science was conducted using a combination of three categories of search keywords, namely:

1. Impairment category: A range of keywords related to older adults and disabilities was used, e.g. {physical” OR “mobility” OR “wheeled mobility device”} OR {“visual” OR “blind” OR “loss of vision”} OR {“hearing” OR “deaf”} OR {“cognitive”} {“elderly” OR “older adults”} OR {“extreme weight” OR “extreme size” OR “obesity” OR “bariatric”} AND

2. Public transportation mode: {“public transit” OR “bus” OR “shuttle” OR “driverless vehicle” OR “autonomous vehicle”} AND

3. Travel chain task: {“station finding” OR “vehicle identification” OR “arrival information”}OR {“boarding/alighting” OR “ingress/egress” OR “entering/exiting” OR “dis/embarking”} OR {“ramp” OR “step” OR “gap” OR “low floor”} OR {“interior circulation” OR “accessible seating” OR “clear floor space”} OR {“securement” OR “wheelchair” OR “fare payment” OR “ride” OR “identify stop”}

Full-text documents such as journal and conference articles, book chapters, and government reports were retrieved and screened by the research team for inclusion in the review if the document:

• addressed one or more travel chain task in the context of accessibility or usability barriers, or accessible and inclusive design,

• referred to at least one impairment category in detail, i.e., accessibility in terms of geographical location or station planning are excluded, and

• investigated a mode of road public transportation that uses shared vehicles, i.e. buses, shuttles, or autonomous vehicles.

Onboard interior circulation was selected for review in this paper due to its influence on safe, independent, and efficient use of vehicle interiors by older adults and people with disabilities, as well as on operational factors such as dwell time, i.e., the time required for serving passengers at a bus-stop. Interior circulation includes sub-tasks pertaining to moving to/from the seat or wheelchair securement location in the vehicle interior that occur after vehicle boarding and prior to vehicle alighting. Studies pertaining to moving to the pick-up location (prior to identifying the station or bus stop), travel from the arrival station to the final destination, wheelchair securement use and design, and rider satisfaction were excluded.

Documents meeting the inclusion criteria were reviewed and the following information extracted: type of document, document title, author(s), year of publication, study methods used, participant sample size(s), main results, design recommendations for accessibility, and conclusions.

RESULTS

A total of 66 documents were identified for inclusion in the overall review. At the time of writing this paper, no human subject study within the literature review directly addressed AVs or LSAS. Results from the review are available in the form of an interactive repository for transportation accessibility research on an open access webpage managed by the University of Michigan’ Inclusive Mobility Research Lab (IMRL, 2019).

Table 1 summarizes the information extracted from the 11 of 66 (17%) reviewed documents related to interior circulation. Eight of the 11 documents focused on users of wheeled mobility devices as the target sample; two studied users with a physical mobility impairment using ambulation aids (e.g., walkers, canes), and two involved users with vision impairments (e.g., using a white cane or service animal). Five of the 11 documents investigated interior circulation in addition to boarding and alighting of transit buses using a full-scale vehicle mock-up in the laboratory (e.g., D’Souza et al. 2017a; 2017b). Other methods used included the secondary analysis of empirical data (e.g., anthropometry data by Bharathy and D’Souza, 2018), and naturalistic observations of in-service transit vehicles (e.g., Hwangbo et al. 2015).

Table 1:

The 11 documents included in the literature review regarding the interior circulation of the passenger travel chain and the design parameters discussed. Some parameters refer to travel chain tasks outside the scope of this paper.

Author, Year	Study Method / Approach	Vehicle Parameter(s) Studied	Impairment Category & Sample Size	Key Conclusions related to Vehicle Interior Design
Bareria et al., 2012	Lab study with transit vehicle mock-up; descriptive statistics	Floor plan configuration	Visually impaired: with mobility cane, n=12 with service animal, n=6 without visual impairment, n=17	Increase legroom between forward-facing seats Design for reduction of circulation time may be the best option for improving passenger experience and reducing dwell time
Bharathy & D’Souza, 2018	Secondary data analysis, Interactive web-based design tool	Clear floor space dimensions	Manual wheelchair, n=277 Powered wheelchair, n=189 Scooter, n=34	Increase the standard for minimum clear floor space Open access web-based design tool to determine dimensions for clear floor space and percent accommodated is introduced
D’Souza, 2014	Lab study with transit vehicle mock-up; inferential statistics	Floor plan configuration, Boarding/Alighting Time, Crowding	Manual wheelchair, n=18 Powered wheelchair, n=21 Scooter, n=9 Visual impairment, n=17 Ambulation aids = 22 No impairment, n=17	Floor plan configurations, social factors, impairment type, and wheeled mobility device type affect the time needed for boarding/alighting and on-board circulation
D’Souza et al., 2012	Lab study with transit vehicle mock-up; descriptive statistics	Boarding/Alighting, Floor plan configuration	Walking aid, n=41	Provide assistive support features (e.g., handrails, stanchions) Increase legroom between forward-facing seats for ease in sitting/rising Increase space for using wheelchairs and storing ambulation aids
D’Souza et al.2017a	Lab study with transit vehicle mock-up; inferential statistics	Low-floor transit buses Boarding/Alighting, Floor plan configuration, passenger load (crowding)	Manual wheelchair, n=18 Powered wheelchair, n=21 Scooter, n=9	Scooter users reported the most difficulty with interior circulation, followed by users of powered and manual wheelchairs. Vehicle interior configuration influenced self-reported difficulty, however no one configuration worked best for all 3 user groups.
D’Souza et al., 2017b	Lab study with transit vehicle mock-up; inferential statistics	Floor plan configuration, Boarding/Alighting time, Passenger load (crowding), Fare payment,	Manual wheelchair, n=18 Powered wheelchair, n=21 Scooter, n=9	Size and location of wheelchair securement area relative to access ramp influence boarding/alighting times Access ramps placed in the middle rather than the front of the bus reduced boarding/alighting time Scooter users required the most time for on-board maneuvering
D’Souza et al., 2010	Secondary data analysis	Clear floor space dimensions	Manual wheelchair, n=195 Powered wheelchair, n=146 Scooter, n=28	Increase the standard for minimum clear floor space powered wheelchairs and scooters are larger and require more floor space compared to manual wheelchairs
Hwangbo et al., 2015	Naturalistic video observations; descriptive statistics	Floor plan configuration	999 incidents analyzed Passengers observed: Elderly, n=200 Physical disability, n=4 Children, n=28 Without physical limitation, n=767	Avoid fragmentary design of space to allow for maneuverability Provide assistive support features
Lenker et al., 2017	Document analysis	Access ramp & gradient, Floor plan configuration, Fare payment, Wheelchair securement	Wheelchairs, N/A	Access ramp gradients below 1:8 (7.1°) are recommended Access ramps located in the middle rather than front of the bus helps reduce boarding/alighting time Longer wheelchair securement space improves maneuvering in/out of the securement area
Robert et al., 2015	Analytical framework for risk calculation	Access ramp gradient, Floor space	Analysis performed for: Powered wheelchair, n=2 Scooter, n=2 Recumbent tricycle, n=2	Consider objective constraints of assistive mobility devices when designing interior vehicle properties, such as securement spaces
Steinfeld et al., 2010	Secondary data analysis	Clear floor space dimensions	Manual wheelchair, n=193 Power chairs, n=146 Scooters, n=28	Increase the standard for minimum clear floor areas Consider the range of wheeled mobility devices that may be used; powered wheelchairs and scooters require greater clear floor space

Collectively, all 11 studies provided evidence of some aspect of increased difficulty with onboard circulation (e.g., maneuvering clearance space for wheeled mobility, problematic seating configurations and seat locations; inadequate assistive features for moving to/from a seat). For instance, onboard circulation space for wheeled mobility device users was found to be inadequate when maneuvering into the device securement space which has direct implications for physical effort, occupant and passenger safety, and bus-stop dwell time (e.g., D’Souza et al. 2017b). The two studies investigating the experiences of persons with vision impairment reported usability issues from a lack of standardized seating configurations, aisle clearances, and legroom between the seats for stowing assistive devices and ambulation aids (Bareria et al. 2012; D’Souza, 2014). Three studies specifically addressed the increased difficulty in circulation due to crowding (D’Souza, 2014; D’Souza et al. 2017a; 2017b), which is an aspect that accessibility standards documents do not address.

Three documents discussed data-driven design tools related to spatial dimensions of vehicle interior design, specifically the clear floor space required for positioning a wheeled mobility device onboard a transit vehicle (Bharathy and D’Souza, 2018; D’Souza et al., 2010; Steinfeld et al., 2010). All three studies found that the current federal accessibility design standards (e.g., US Access Board, 2010) for clear floor space dimensions are insufficient for accommodating the full range of contemporary wheelchairs and scooters. Bharathy and D’Souza (2018) also provided an online design tool for calculating clear floor space dimensions to accommodate a desired proportion of wheelchair users.

DISCUSSION AND CONCLUSIONS

Realizing the potential of LSAS to address the mobility needs of people with disabilities and older adults will require that diverse users’ needs are understood early in design, and that emerging LSAS technologies are then designed to meet these needs. This review study found that accessibility research on shared-use LSAS was non-existent, while accessibility research on other public transit modes was limited. The latter was mostly focused on physical mobility impairments, however continued research is needed to overcome user-specific challenges and barriers across the transportation travel chain (Audirac, 2008; Risser, Iwarsson, & Ståhl, 2012).

Designing for safe and efficient onboard circulation for passengers of diverse mobility and sensory abilities is an important aspect of accessibility. The documents reviewed consistently reported an increased difficulty and/or increased dwell time associated with onboard circulation for passengers with mobility impairments in public transit vehicles, suggesting inadequacies in current vehicle designs.

The two studies on sensory impairment were related to visual impairment; neither addressed auditory impairment even though people with auditory impairment experience barriers in information access that lead to delays or errors when traveling (Bettger, 1989; Fürst & Vogelauer, 2012). Older adults and people with limitations in cognitive function have also been shown to experience challenges in public transportation (Broome et al., 2009). Currently, many of the onboard issues for people with sensory and cognitive impairments can be addressed with help from the bus driver (e.g., availability of a vacant seat, expected arrival time to the destination stop). Driverless LSAS will increase the significance of providing onboard information to such passengers. Future accessibility research on LSAS and automated vehicles will need to adopt an inclusive design approach by focusing on multiple impairment groups to understand common needs and potential tradeoffs across the range of physical, sensory and cognitive abilities.

The current review also highlights the need for better design of information (i.e., design tools and resources) to support the physical design of transportation vehicles. The interior configuration of vehicles (clear floor space, seating configurations, aisle spaces) had a profound effect on the degree of difficulty experienced by passengers with disabilities, even when the vehicle interiors were compliant with federal accessibility standards (US Access Board, 2008). The accessible design of vehicle interiors will be a key design consideration for supporting safe, independent and efficient use of emerging LSAS technology by users with mobility impairments. However, design tools and information to support accessible design for these user groups are few, as evidenced by the current review.

One method for studying the accessibility of future or concept LSAS designs prior to development and deployment is to use full-scale, reconfigurable mock-ups in a laboratory setting. Five of the 6 studies reviewed involving human subjects made effective use of a static reconfigurable transit vehicle mock-up in a laboratory environment. This method allowed researchers to engage end-users with disabilities in co-investigating accessibility features and vehicle interior configurations. Going forward, user studies with mock-ups could help to proactively investigate accessibility issues and generate actionable information to support accessible and inclusive design of LSAS.

A limitation of the literature review performed was the exclusion of some travel chain components such as moving to and from transportation stations and stops, i.e., first and last mile concerns. The design of transportation facilities, sidewalks, and street crossings are also important to consider when designing for an accessible transit chain; however, these issues are often ignored (Hwangbo, Kim, Kim, & Ji, 2015; Risser et al., 2012). Extending the review to include these stages of the travel chain may provide policymakers and transportation planners with more complete information when planning for LSAS deployments.

The preliminary work presented in this paper is part of a larger research effort at the Inclusive Mobility Research Lab to develop a searchable web-based repository for current and future research in accessible transportation (IMRL, 2019). The repository will provide a centralized and categorized database intended to help accessibility researchers, policymakers and industry to review previous accessibility research, and identify knowledge gaps and directions for future accessibility research related to automated shuttles.