Investigating Inclusive Design of Shared Automated Vehicles with Full-scale Modeling

Abstract

Shared automated vehicles (SAVs) in the form of low-speed driverless shuttles have the potential to improve independent mobility for older adults and people with disabilities. At full vehicle autonomy and in the absence of an onboard operator, tasks such as ingress-egress, interior circulation, and securement of passengers and carry-on items will need to be safe, efficient, and independent. This paper describes a novel laboratory apparatus for conducting inclusive design research related to SAVs and presents preliminary findings from an ongoing preliminary study examining the effects of interior design configuration on ingress-egress performance for six wheelchair users. Early findings emphasize the interactions between diverse user abilities and technology design on user performance. The study demonstrates the potential benefit of full-scale physical simulations to investigating a broad range of usability and inclusive design issues related to emerging SAVs.

Full text

INVESTIGATING INCLUSIVE DESIGN OF SHARED AUTOMATED
VEHICLES WITH FULL-SCALE MODELING
Kamolnat Tabattanon, Patrik T. Schuler, Clive D’Souza
Department of Industrial & Operations Engineering, University of Michigan, Ann Arbor, MI
Abstract
Shared automated vehicles (SAVs) in the form of low-speed driverless shuttles have the potential
to improve independent mobility for older adults and people with disabilities. At full vehicle
autonomy and in the absence of an onboard operator, tasks such as ingress-egress, interior
circulation, and securement of passengers and carry-on items will need to be safe, efficient, and
independent. This paper describes a novel laboratory apparatus for conducting inclusive design
research related to SAVs and presents preliminary findings from an ongoing preliminary study
examining the effects of interior design configuration on ingress-egress performance for six
wheelchair users. Early findings emphasize the interactions between diverse user abilities and
technology design on user performance. The study demonstrates the potential benefit of full-scale
physical simulations to investigating a broad range of usability and inclusive design issues related
to emerging SAVs.
INTRODUCTION
Shared automated vehicles (SAVs) have gained much attention from transportation service
providers, federal transportation agencies, vehicle manufacturers, and the media as an
emerging transportation mode for addressing long-standing challenges in first-/last-mile
mobility. Multiple companies have already started testing and/or deploying SAVs in the US
on a limited basis and/or on specific routes (Cregger et al. 2018). Early-use cases suggest
that SAVs could substantially increase independent mobility for consumers with disabilities,
older adults, and people that are otherwise ineligible or unable to drive (Claypool et al.,
2017).
Despite its potential, in reality most SAVs are not designed with accessibility in mind.
Usability and accessibility of SAVs includes the ease of access to physical interactions and
communication with vehicles, including information access and reliable plans for emergency
action (NCMM, 2018). At present, no federal or industry guidelines or design aids for
accessibility specific to driverless SAVs exist. Guidance regarding accessible design of SAVs
is the result of interpreting existing standards by DOT, NHTSA, and the US Access Board
for conventional buses and vans. The lack of inclusive SAV design and deployment is
evident from multiple high-profile reports that highlight the urgent need for research to
identify inclusive design requirements, develop technology interventions, and create
methods to evaluate the usability of SAVs (Claypool et al., 2017; NCD, 2015; ITSA et al.,
2019).
HHS Public Access
Author manuscript
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March
23.
Published in final edited form as:
Proc Hum Factors Ergon Soc Annu Meet
. 2021 February ; 64(1): 965–969.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Harnessing the full potential of SAVs to provide inclusive and equitable service for
consumers of diverse abilities will require an evidence-base for developing standards, setting
policy, and making design decisions. The existing human factors literature lacks a systematic
empirical examination of inclusive design needs specific to the usability of SAVs
(Tabattanon et al., 2019). This includes issues pertaining to spatial needs for manageable
access ramp gradients, ingress-egress doorway locations, interior circulation, seating
configurations, wheeled mobility securement locations, and in-vehicle communication and
information access (Tabattanon et al., 2019; Claypool et al., 2017; ITSA et al., 2019).
Ideally, the research to support such an evidence-base is needed prior to mass deployment of
SAVs in order to minimize accessibility-related retrofitting and after-market adaptations
which tend to be costly and sub-optimal and, further, could potentially compromise vehicle
performance (Rojas et al., 2020).
A promising methodology in participatory human factors research is the use of laboratory-
based, full-scale physical mock-ups (Steinfeld, 2004; Steinfeld et al., 2010). Previous
laboratory studies on accessible transit buses (e.g., Bareria et al., 2012; D’Souza et al., 2017;
2019) and passenger rail (e.g., Daamen et al., 2008; Tabattanon and Hunter-Zaworski, 2018)
have successfully used full-scale mock-ups to engage users in generative studies to
understand needs and perspectives of diverse consumers and in evaluative studies to quantify
effects of different concept designs on task performance and usability.
The objective of this paper is to introduce a methodology employing a full-scale
reconfigurable mock-up of an SAV for engaging consumers with disabilities and older adults
in human factors research related to inclusive SAV design. Preliminary data from an ongoing
study with manual and powered wheelchair users is presented to illustrate the proposed
methodology and its potential for understanding user needs and generating actionable design
information.
METHODS
Full-scale Mock-up Apparatus
Based on a review of existing SAVs (e.g., Navya Autonom, EasyMile-10, Local Motors
Olli), a full-scale reconfigurable mock-up of an SAV was constructed in the laboratory
(Figure 1). The outer sides of the mock-up can be adjusted to simulate different form-factors
of SAVs up to 2m width × 4m length. The height-adjustable vehicle platform (between 10cm
to 50cm) allows for different step heights and ramp slopes at ingress-egress. A folding
access ramp (76cm width × 122cm) attached to the curbside frame can achieve ramp
gradients ranging between 0° (level boarding) to 14° degrees (1:4 slope). Movable vertical
posts are used to simulate a doorway along the curbside of the vehicle. Repositionable grab-
bars can be fastened vertically on the door frame or along any length of the mock-up frame
to introduce assistive support features.
The vehicle mock-up has up to 14 moveable seats and a wheeled mobility securement
system (Q’Straint Quantum) mounted on a moveable base which allows the simulation of
different interior seating configurations and wheelchair securement locations. The seats can
be positioned facing inward along the front, rear, and sides of the vehicle interior to simulate
Tabattanon et al. Page 2
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

different seating capacities and configurations of perimeter seating. Available clear floor
space under each seat can be reduced by placing blocks underneath, thus simulating
situations where electrical or drive mechanisms (e.g., electric batteries, AV sensing and data
storage equipment) are packaged beneath the seats to reduce the overall size and weight of
the SAV. Likewise, movable wooden blocks are used to simulate wheel-wells.
The wheelchair securement system used is an electromechanical semi-automated system that
does not involve tie-down straps (Perez et al., 2019). The system is intended for use in a
rear-facing orientation. To operate the system, the wheeled mobility occupant maneuvers
into the securement space to a rear-facing position, centers their wheelchair on the wheeled
mobility securement area, and presses a one-touch button that initiates the securement
process. Securement is achieved by means of two horizontal arms – one adjacent and
parallel to the vehicle wall and the other a rotating arm mounted on the aisle-side of the
securement area. When activated, the aisle-side arm telescopes out from the base unit to
clear the passenger’s wheelchair and rotates from a vertical position to a horizontal position.
Both arms then move inward to squeeze the wheelchair with a 50lbs force. Wheelchair users
press the same button to initiate release of the securement mechanism.
Instrumentation
The laboratory space is instrumented with an optical motion capture system and multiple
video cameras. The frame of the mock-up apparatus allows for recording movements of
participants using motion capture and video of the vehicle interior during ingress, interior
circulation, securement, and egress. The apparatus also provides the opportunity for
introducing prototypes of new equipment such as communication devices for user evaluation
(e.g., interactive touch screens, in-vehicle information displays).
Study Overview and Participants
In order to demonstrate the utility of this full-scale mock-up for human factors research, a
preliminary study was conducted to examine the effects of a few key design features on
ingress-egress performance of wheelchair users in an SAV with an 11-passenger perimeter
seating configuration and one wheeled mobility securement area (WMSA). The study
recruited participants from the local community who use either a manual or powered
wheelchair as their primary means of mobility. The study protocol was approved by the
university’s institutional review board.
The study manipulated two variables: (1) available floor space (low: without clearance space
beneath the forward and rear seats, vs. high: with the entire space beneath the seats available
as clear floor space; Figure 2) and (2) curb-side doorway location (centered vs. rearward).
Ingress-egress task completion times obtained from a video-based task analysis and a
qualitative analysis of participant feedback comments to probing design questions were used
to operationalize task performance.
Study Procedure
Initially, a questionnaire was administered to obtain information about participant
demographics (e.g., sex, age, years using a mobility device), medical condition (i.e., reasons
Tabattanon et al. Page 3
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

for using a mobility device), and anthropometry (e.g., body mass, occupied device mass,
occupied and unoccupied device depth and width, knee height). A functional test of device
maneuverability was performed to obtain the minimum length of a rectangular space needed
for completing a parallel park maneuver using a method of limits starting at their occupied
depth and incrementing by 2.5cm (~1in.) each iteration until successful.
Simulated Ingress-Egress Trials
The experiment procedure had participants independently perform three repetitions each of
four timed ingress-egress trials (two levels of floor space × two door locations) on the SAV
mock-up. The general sequence of steps for
ingress
consisted of ramp ascent, maneuvering
to the WMSA, backing into and positioning in the WMSA in a forward-facing direction, and
reaching to and pressing a one-touch button to initiate the device securement process. The
procedure for
egress
consisted of reaching to and pressing the button to initiate the
securement release, maneuvering out of the WMSA toward the doorway followed by ramp
descent. For this study, the access ramp was maintained at a constant slope of 1:12 to
minimize fatigue.
All of the ingress-egress trials were recorded using two synchronized video cameras. One
camera was positioned by the access ramp to capture ramp ascent/descent. A second video
camera was placed near the front of the vehicle and directed rearward to record movements
associated with interior circulation. During the third repetition of each configuration, a
researcher approached the participant immediately after the securement process was
completed and asked open-ended questions about the interior configuration and potentially
traveling in the current position. Participants also provided post-trial comments following
each configuration.
Dependent Measures
This paper reports on preliminary findings from six participants. For each trial, ingress-
egress task times and task completion strategies employed by participants were extracted
using video data from the two cameras.
Ingress time
was defined as the duration from the
instant when the participant or wheelchair first contacts the access ramp for ramp ascent
until the instant when the participant is positioned in the wheeled mobility securement area
and activates the device securement button.
Egress time
was defined as the time starting
from pressing the button to initiate securement release through maneuvering and ramp
descent until to wheelchair is no longer in contact with the ramp.
Due to the limited sample size, analyses were limited to descriptive analysis (and not
inferential statistical analysis). Ingress and egress times from all three repetitions were
averaged per participant and plotted by device type, available floor space, and door location.
Open-ended responses from participants were coded and summarized by theme.
Tabattanon et al. Page 4
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

RESULTS
Participant Characteristics
Table 1 summarizes the key demographic and anthropometric characteristics of the study
sample. On average, the occupied powered wheelchairs weighed more, had a longer
occupied length, and required more space for performing a parallel park maneuver. Of the
six participants, four reported using their mobility device for over eight years; while one
manual and one powered wheelchair user each reported using their device for between one-
three years.
Ingress and Egress Task Times
Figure 3 depicts the ingress and egress task times for each participant (averaged over three
repetitions) by available floor space and door location. Overall, ingress-egress times differed
substantially between participants compared to differences between design conditions. This
highlights the heterogeneity in functional abilities among populations with disabilities and
underscores the challenges of and need for conducting inclusive design research.
Patterns in task times need to be interpreted with some caution. It appears that task times for
powered wheelchair users were slightly shorter compared to times for manual wheelchair
users, except for ingress at low levels of available floor space with a centered door location.
In this configuration, two powered wheelchair users had long ingress times while one
participant consistently had times ≤13sec. for both ingress and egress.
Video analysis revealed a range of maneuvering paths and strategies being employed by
users of powered wheelchair compared to users of manual wheelchairs. Two powered
wheelchair users chose to ascend the ramp facing rearwards however they did not employ
this approach in all trials for the same configuration. All three manual wheelchair users
ascended the ramp facing forwards and generally repeated a similar interior circulation path
across repetitions in any one configuration.
Participant Feedback Comments
Responses to open-ended questions emphasized design barriers to safe and efficient interior
circulation. Configurations without the available space beneath seats evoked negative
sentiments, e.g., “
this is not enough spac
e” and “
my feet hit the boxes
”, while high levels of
available floor space evoked positive sentiments, e.g., “
this is better because there is more
room
” and “
there was better room to maneuver
.” Two participants commented on the benefit
of flip-up seats, e.g., “
flipping up seats gave me more control of the path to the securement
”.
Feedback comments also yielded insight into decisions, e.g., choice of maneuvering path,
akin to contextual inquiries. For example, one powered wheelchair user explained their
decision to enter the vehicle facing rearwards because it was “
equally comfortable to back
in
”, even in real-world settings, which made it “
easier and faster to just turn into the
securement
”. Some comments related to experiences on conventional transit vehicles
suggested there are unmet needs for accessible features; for example, one participant gave
suggestions for positioning grab-bars and buttons to aid with reach and locating information
Tabattanon et al. Page 5
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

displays for improved line-of-sight. A subset of comments addressed psychosocial factors,
e.g., “
I wouldn’t like to be facing people
” and “
I could move into the space, but if there were
people here, they would have needed to move
” – which highlights another benefit of full-
scale environmental mockups, namely, the ability to provide situational context and help
elicit past experiences.
DISCUSSION AND CONCLUSIONS
This paper illustrates the benefits of using environmental mock-ups in the context of
inclusive design research on emerging SAVs. A key benefit demonstrated is the ability to
simulate and study effects of multiple design options within a controlled environment. Such
experimental control is not possible in field-based studies with operational vehicles. While
video-based analyses are possible within observational studies of in-service transit vehicles
(e.g., Frost et al. 2015; Hwangbo et al., 2015; Kostyniuk and D’Souza, 2020) manipulating
features such as door locations and floor space is often not feasible. The few SAVs presently
in-service make naturalistic field studies challenging.
While this was an initial study with a limited sample, our early findings suggest an
interaction between users’ functional abilities and environmental affordances on usability
and task performance. In other words, the magnitude of the effect of different design
configurations differed by participant. Further, a study with a larger sample size and more
detailed measurements is needed to examine person-environmental relationships of practical
significance.
Use of full-scale environmental simulations in human factors experiments also lends support
to mixed-methods analyses. Besides quantitative measures of task times, this study also
conducted extensive interviews with participants while onboard the vehicle and after they
had exited the vehicle. Unlike field-based studies (e.g., think-aloud protocols on in-service
transit vehicles; Risser et al., 2012), participants in our study were not pressed for time nor
distracted by external uncontrollable factors (e.g., traffic noise, inclement weather).
Participants received time to process the questions and give thoughtful responses while in
the simulated environment. Further, it was observed that participants continued to cross-
reference and/or gesture to features in the mock-up when providing feedback comments
such as when describing difficult movements/maneuvers or providing suggestions for design
improvement.
Inclusive design by definition requires consideration to user with mobility, sensory,
communication, or cognitive impairments. Presently, this study only recruited wheeled
mobility device users. However, the methodology using environmental simulations can be
extended to a diverse spectrum of consumers that could benefit from emerging driverless
SAVs (e.g., blind and low vision consumers, older adults) – and is the topic of an ongoing
study.
In conclusion, preliminary findings support the use of reconfigurable physical simulations to
study a broad spectrum of usability and inclusive design issues related to emerging SAVs. A
range of test configurations and concept designs could be simulated in a short span of time.
Tabattanon et al. Page 6
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Multiple human factors and ergonomics research techniques can be leveraged to obtain
information regarding physical constraints, person-environment interaction, decision-
making, and performance. This study demonstrates the potential benefits from a relatively
small investment into full-scale physical simulations to investigate usability issues and
inclusive design considerations in a proactive manner early in design and/or prior to system
deployment.
ACKNOWLEDGEMENT
The contents of this paper were developed under a grant from Mcity, an office within the University of Michigan
Office of Research, and from the National Institute on Disability, Independent Living, and Rehabilitation Research
(NIDILRR grants #90IF0094 and #90RTHF0001). NIDILRR is a Center within the Administration for Community
Living (ACL), Department of Health and Human Services (HHS). The contents of this paper do not necessarily
represent the policy of nor endorsement by NIDILRR, ACL, HHS, or the Federal Government.
The authors thank student assistants Kip A. Schimmoeller, Kieran T. O’Brien, and Baibhav Kumar Panda for their
assistance with data collection.
REFERENCES
Bareria P, D’Souza C, Lenker J, Paquet V, & Steinfeld E (2012). Performance of Visually Impaired
Users during Simulated Boarding and Alighting on Low-Floor Buses. Proceedings of the 56th
HFES Annual Meeting, pp: 656–660. DOI: 10.1177/1071181312561137
Claypool H, Bin-Nun A and Gerlach J (2017). Self-driving cars: The impact on people with disabilities
[White paper]. Ruderman Family Foundation, Newton, MA.
Cregger J, Dawes M, Fischer S, Lowenthal C, Machek E, & Perlman D (2018). Low-Speed Automated
Shuttles: State of the Practice (Report No. FHWA-JPO-18–692). John A Volpe National
Transportation Systems Center, USDOT.
Daamen W, Lee Y.-c. & Wiggenraad P (2008). Boarding and Alighting Experiments: Overview of
Setup and Performance and Some Preliminary Results. Transportation Research Record, 2042, pp:
71–81. DOI: 10.3141/2042-08
D’Souza C, Paquet V, Lenker J, & Steinfeld E (2019). Self-reported difficulty and preferences of
wheeled mobility device users for simulated low-floor bus boarding, interior circulation and
disembarking. Disability & Rehabilitation: Assistive Technology, 14(2), 109–121. DOI:
10.1080/17483107.2017.1401128 [PubMed: 29130752]
D’Souza C, Paquet V, Lenker J & Steinfeld E (2017). Effects of Transit Bus Interior Configuration on
Performance of Wheeled Mobility Users during Simulated Boarding and Disembarking. Applied
Ergonomics, 62, 94–106. DOI: 10.1016/j.apergo.2017.02.008. [PubMed: 28411744]
Frost KL, Bertocci G, & Smalley C (2015). Ramp-related incidents involving wheeled mobility device
users during transit bus boarding/alighting. Archives of physical medicine and rehabilitation, 96(5),
928–933. DOI: 10.1016/j.apmr.2014.12.013 [PubMed: 25576087]
Hwangbo H, Kim J, Kim S, & Ji YG (2015). Toward universal design in public transportation systems:
An analysis of low-floor bus passenger behavior with video observations. Human Factors and
Ergonomics in Manufacturing & Service Industries, 25(2), 183–197. DOI: 10.1002/hfm.20537
ITSA, Bayless S, and Davidson S (2019). Driverless Cars and Accessibility: Designing the future of
transportation for people with disabilities. The Intelligent Transportation Society of America (ITS
America), 4 2019.
Kostyniuk L & D’Souza C (2020). Effect of Encumbrance and Mobility Aid Use on Dwell Time in
Low-floor Transit Vehicles. Transportation Research Part A: Policy and Practice, 132 (Feb 2020),
872 – 881. DOI: 10.1016/j.tra.2020.01.002.
NCMM (2018) Autonomous Vehicles: Considerations for People with Disabilities and Older Adults.
National Center for Mobility Management (NCMM). Retrieved from: https://bit.ly/2TDNLeg
NCD (2015). Self-Driving Cars: Mapping Access to a Technology Revolution (Publication No.
01656924). National Council on Disability, Washington, DC.
Tabattanon et al. Page 7
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Perez B, Choi J, Paquet V, Lenker J, Kocher L, Nemade M, Kern C, and Steinfeld E (2019).
Comparison of wheelchair securement systems designed for use in large accessible transit vehicles
(LATVs). Assistive Technology, 2019, 1–11. DOI: 10.1080/10400435.2019.1604582
Rojas JF, Tabattanon K, Goberville N, Sandhu N, Kiel M, D’Souza C and Asher ZD (2020). A Case-
Study Evaluating the Performance of an Autonomous Electric Shuttle Modified for Wheelchair
Access. Proceedings of the Transportation Research Board Annual Meeting, Washington, DC.
Risser R, Iwarsson S, & Ståhl A (2012). How do people with cognitive functional limitations post-
stroke manage the use of buses in local public transport? Transportation research part F: Traffic
Psychology and Behaviour, 15(2), 111–118. DOI: 10.1016/j.trf.2011.11.010
Steinfeld E (2004). Modeling spatial interaction through full-scale modeling. International Journal of
Industrial Ergonomics, 33(3), 265–278. DOI: 10.1016/j.ergon.2003.06.004
Steinfeld E, Paquet V, Lenker J, D’Souza C & Maisel J (2010). Universal Design Research on
Boarding and Using Buses. Proceedings of the 12th International Conference on Mobility and
Transport for Elderly and Disabled Persons (TRANSED), Hong Kong, 6 2010.
Tabattanon K, Sandhu N & D’Souza C (2019). Accessible Design of Low-Speed Automated Shuttles:
A Brief Review of Lessons Learned from Public Transit. Proceedings of the 63rd HFES Annual
Meeting, pp: 526 – 530. DOI: 10.1177/1071181319631362.
Tabattanon K, & Hunter-Zaworski K (2017). Accessible Sleeper for the Next Generation Passenger
Rail Vehicles. Proceedings of the 2018 Joint Rail Conference, American Society of Mechanical
Engineers Digital Collection (JRC2018-6104, V001T08A001). DOI: 10.1115/JRC2018-6104
Tabattanon et al. Page 8
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Figure 1:
Image showing the experimental apparatus. The configuration depicted contains 11 seats,
one automatic securement device, and a centered doorway
Tabattanon et al. Page 9
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Figure 2:
Images depicting low (left) and high (right) levels of available floor space achieved by
manipulating the clearance space beneath the seats.
Tabattanon et al. Page 10
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Figure 3:
Ingress times (top panel) and egress times (bottom panel) by door location (left vs. right
panels) and available floor space averaged over three repetitions for each of the six
participants, i.e., three manual wheelchair users (solid lines), and three powered wheelchair
users (dotted lines).
Tabattanon et al. Page 11
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript

Author Manuscript Author Manuscript Author Manuscript Author Manuscript
Tabattanon et al. Page 12
Table 1:
Participant Characteristics (N = 6)
Manual (n = 3) Powered (n = 3)
Demographics:
Gender: Male, Female 1, 2 2, 1
Age, Mean (SD) years 40.7 (11.7) 41.7 (18.5)
Anthropometry:
Occupied Mass (kg) 96.2 (21.6) 235.7 (16.1)
Occupied Depth (cm) 93.3 (8.5) 115.7 (2.1)
Occupied Width (cm) 76.7 (12.8) 71.8 (6.3)
Knee height (cm) 61.7 (6.8) 65.6 (1.9)
Parallel-park Dist.(cm) 105 (0.0) 119.2 (2.5)
Proc Hum Factors Ergon Soc Annu Meet
. Author manuscript; available in PMC 2021 March 23.