Passenger comfort and trust on first-time use of a shared autonomous shuttle vehicle

https://doi.org/10.1016/j.trc.2020.02.026Get rights and content

Highlights

  • Novel contribution analysing user trust and comfort in a shared autonomous vehicle.

  • Statistically significant relation between trust and speed and direction of face.

  • Strong correlation between comfort and trust.

  • Car drivers in particular showed increased favourability after the experience.

Abstract

Autonomous Vehicles (AV) may become widely diffused as a road transport technology around the world. However, two conditions of successful adoption of AVs are that they must be synchronously shared, to avoid negative transport network and environmental consequences, and that high levels of public acceptance of the technology must exist. The implications of these two conditions are that travellers must accept sharing rides with unfamiliar others in Shared Autonomous Vehicles (SAV). Two factors that have been identified as being positive influencers of acceptance are comfort and trust. The present paper undertakes a novel examination as to how comfort and trust ratings are affected by specific attributes of the ride experience of travelling in a fully-automated real-world, shared vehicle. To this end, 55 participants experienced riding in an SAV shuttle under experimental conditions at a test facility. Each experimental run involved two unrelated participants, accompanied by a safety operative and a researcher, undertaking four trips in the SAV, during which two conditions were presented for each of the independent variables of ‘direction of face’ (forwards/backwards) and ‘maximum vehicle speed’ (8/16 km/h). Order of presentation was varied between pairs of participants. After each run, participants rated the dependent variables ‘trust’ and ‘comfort’ (the latter variable comprised by six comfort factors). Expected and evaluative ratings were also obtained during pre-experimental orientation and debriefing sessions. Statistically significant relationships (p < .001) were found between trust and each of the independent variables, but for neither variable in the case of perceived comfort. A strong correlation was found between comfort and trust, interpreted as indicating trust in the SAV as an important predictor of perceived comfort. The before and after-experiment ratings for both variables showed statistically significant increases, and particularly for daily car drivers.

Introduction

Predicting the market penetration of automation technologies into the road transport sector is complex. Applying a scenario analysis approach, Lyons and Babbar (2017) identified a range of 8–84% for highly and fully automated vehicles as a share of global new vehicle sales in 2035, with a ‘central case’ of 25%, assuming rapid technological development and moderate global uptake. More recently, analysing the North American context, Litman (2017) predicts around a 35% share of sales by the same year, with around one-fifth fleet penetration, but accelerating to 50% of fleet composition around 2050. Given the potential scale of change, it is essential that policymakers are able to respond drawing upon a strong evidence base about likely effects and consequences.

Whilst benefits from automation, notably relating to improved road safety, are widely promoted, the prospect of a rapid transition occurring within a laissez-faire policy framework has prompted major concerns about the possible negative consequences of rising demand due to falling user costs. Where a fare is paid, these are expected to fall if driver costs are eliminated. Similarly, travel time costs, whether a real business cost or a perceived personal cost, would be expected to fall if people who currently self-drive are able to reinvest travel time on more productive activities (Diels et al., 2017). Simulation demand modelling of such future market contexts has indicated that traffic and congestion in a typical city could double due to cost reductions (International Transport Forum - ITF, 2015), as could energy consumption and greenhouse gas emissions (Wadud et al., 2016). However, many commentators identify that a significant level of shared automated mobility could mitigate for the increase in demand (e.g. Fulton et al., 2017, McKinsey, 2016, NACTO, 2016), with the additional personal travel resulting in only modest increases in vehicle traffic and congestion (ITF, 2015), and energy and emissions potentially halved (Wadud et al., 2016).

To achieve this relatively benign adoption scenario for AVs, though, the key condition is that the vehicles must be synchronously shared. This distinction is important as the term ‘shared’ is often vaguely applied in the transport policy discourse. In fact, the assumption present in many visions of ‘future mobility’ that vehicle sharing will be a significant or dominant phenomenon is mainly based on aspiration or opinion (Parkhurst and Lyons, 2018). Also, it is worth noting that the existing, limited, evidence on willingness to share identifies significant social and psychological barriers (Merat et al., 2017). The present paper therefore seeks to contribute to this scarce evidence base through an experimental exploration of factors influencing SAV acceptance.

The term Shared Autonomous Vehicle (SAV) is in current usage referring to both (i) a vehicle exclusively used by individual travel parties and subsequently used exclusively by other parties (asynchronous sharing) and (ii) used by more than one individual/group, each of which accepts sharing space in the vehicle for whole or part of the journey with others, who might be acquainted or strangers (synchronous sharing). In this paper we refer exclusively to the latter definition.

Several authors claim SAVs can provide greater benefits than AVs, especially due to the expected reduced need for parking space and reduced congestion (Liu, 2018). However, these benefits depend on the market penetration of SAVs, which is currently estimated considering simulation that might not be realistic (Narayanan et al., 2020). In principle, SAV services could be offered in niches currently served by human-driven taxis, taxi-buses or buses. Whilst, they would tend to compete or replace taxis and taxi-buses, they might either replace or integrate with established fixed route public transport systems (Levin et al., 2019), in which case they might focus on low-demand routes within a network, or provide feeder services to trunk routes, or service low-demand periods of the day (Shen et al., 2018). In the implementations and demonstrations to date, the term ‘shuttle’ is often applied. The vehicles are typically capable of carrying 4–12 people. However, according to Vosooghi et al. (2019) the financial sustainability of a SAV service is strongly correlated to the fleet size, and the benefits of having more than four seats in SAVs might be limited. Given this relatively low vehicle capacity, the ‘business model’ will ultimately require full automation to Level 4 capability to be viable, implying no handover to a driver would be necessary within a defined spatial environment (Society of Automotive Engineers International – SAE, 2018) and for operation to be on-demand. Hitherto, the demonstrations in environments with public access have been at slow speeds and on constrained routes. Higher-speed operation would need a significant improvement in technology (e.g. sensors, software), adaptation of the infrastructure, and a clear definition of specific regulations (Schreurs and Steuwer, 2016).

SAVs have been trialled within projects in Europe (e.g. CityMobil1 (2006–2011) and CityMobil2 (2012–2016; WePods in the Netherlands (Liang et al., 2016, Van der Wiel, 2017); Smartshuttle in Switzerland (Eden et al., 2017); EUREF in Berlin-Schöneberg (Nordhoff et al., 2018), in the US (First Transit/Easymile), and in Australia (Navya). As the state of the art is currently one of demonstration projects, rather than full, permanent public services, actual adoption cannot be measured, but factors affecting target users’ acceptance have been identified and confirmed. Alessandrini (2016) reported that users of demonstration services in the Citymobil2 project gave high ratings of comfort and safety, although it was observed that the stakes around trust were particularly high. In part this was due to the radical step of replacing the driver with artificial intelligence and robotics, but also due to the strong claims made about the potential of the technology to transform transport systems by improving road safety, traffic efficiency, air quality, and access to mobility services (Alessandrini et al., 2014). Comfort can also be expected to influence the ongoing acceptance and adoption of AVs, and this is particularly important in the case of SAVs, due to the presence of other travellers, which will constrain choices about standing/sitting position in the vehicle, as well as potentially bring unfamiliar others into close physical proximity.

However, comfort and trust are very subjective, and so both human factors and people’s perceptions become important in understanding how to design new vehicles and transport systems to make them more attractive for potential users. SAV services built around a ridesharing model will rely on social, as well as technical, innovation; that there will be a future willingness amongst travellers to share a small vehicle with strangers, despite the absence of a physical presence of an operative ‘in authority’, which might actually render synchronous sharing less attractive than it is now. The present paper therefore contributes to enhancing understanding of the perceptions of first-time users of a SAV providing data about acceptance in the context of a test-track environment with Level 4 operation of a four-seat vehicle. Full details of the experiment are given in Section 2. A particular focus was to examine the interactions between trust and comfort, based on these having been identified as key variables in the literature (Siebert et al., 2013, Bellem et al., 2018).

Whilst the evidence base specifically on SAVs is limited, trust has been widely recognised as an important factor in the acceptance and utilisation of automation across different sectors, as it both depends on people’s beliefs toward automation and influences their intention to use it (Carter and Bélanger, 2005, Choi and Ji, 2015, Körber et al., 2018, Gefen et al., 2003, Lee and Moray, 1992, Lee and Moray, 1994, Lee and See, 2004, McKnight et al., 2002, Merritt and Ilgen, 2008, Noy et al., 2018, Parasuraman et al., 2008, Parasuraman and Riley, 1997, Pavlou, 2003, Shariff et al., 2017, Siebert et al., 2013). According to Du et al. (2019), understanding what kind of factors influence trust in automation is very important for a better understanding of AV use. For this reason, trust has been identified as a key issue for AV acceptance and adoption (Bazilinskyy et al., 2015, Verberne et al., 2012, Bansal et al., 2016, Molnar et al., 2018, Zhang et al., 2019), and according to Morgan et al. (2018), it can be considered “one of the most important enablers (and indeed barriers) to humans adopting and continuing to use new automation technology”. However, trust is a subjective factor and depends on the personality of the individual (e.g. differences in the propensity to trust) and sociocultural context (e.g. social norms and expectations) in which the decision to trust takes place (Lee and See, 2004). Notably in the context of the current paper, Molnar et al. (2018) found that people who prefer being a passenger rather than a driver were more accepting of the concept of AVs.

Khastgir et al. (2018:291) adapted the definition provided by Lee and See (2004), defining trust as “a ‘history-dependent’ attitude that an agent will help achieve an individual’s goals in a situation characterised by uncertainty and vulnerability”. Following Khastgir et al., the inclusion of ‘history-dependent’ in the previous definition highlights the importance of previous knowledge about the system on trust, as unfamiliarity with AV technology might have a negative impact on public acceptance (Dong et al., 2019). According to Du et al. (2019), increasing the level of information about AVs can reduce potential users’ anxiety, increase their trust in AVs and the likelihood that they will exhibit positive attitudes towards AVs. Furthermore, the social learning theory approach assumes that expectations for specific events strongly depend on previous experiences on similar events or situations (Rotter, 1971). This is supported by Gold et al., 2015, Hartwich et al., 2018, and the Venturer Project Partners (2018), who found that participants’ self-reported trust in AVs increased after experiencing automated driving across experimental runs in a simulation trial, and by Xu et al. (2018), who found that direct experience increased participants’ trust, perceived usefulness, and perceived ease of use ratings towards AVs.

Lee and See (2004) stated that trust in automation is strictly related to “emotions on human-technology interaction”, which is a key factor for acceptance, but is also important for safety and performance. For this reason, it should be a factor considered when designing complex, high-consequence systems like AVs. Despite the difference between interpersonal trust and trust in technology, they have some similarities (Hoff and Bashir, 2015). For example, Parasuraman and Riley (1997) suggest people’s trust in technology can be considered as being akin to their degree of trust in the designers of technological systems.

Fig. 1 presents a summary of the main factors influencing trust that have been identified above. It includes the three main groups identified by Hoff and Bashir (e.g. dispositional trust, situational trust, learned trust), factors related to expectations, personality and emotions of individuals, and finally, external factors like social norms and people’s beliefs about AVs. In the context of a SAV, social norms about shared transport-related behaviours are likely to be important, as will be the influence of personality, experience and expectations, and situational and learned trust on an individuals’ willingness to share. Age, gender, and culture can be expected to interact with these factors. In the present study, expectations were assumed to be weakly defined, and learned trust low, due to the rareness of exposure to SAVs to date, whilst social norms were expected to be drawn from other travel experiences, but with the norms and expectations about behaviour in an experimental context also having an influence. Some aspects of experience, beliefs and personality were addressed through survey data, whilst the experiment was primarily a test of situational trust.

Fig. 1 also shows there is a relationship between trust and comfort. Bellem et al., (2018), citing Siebert et al. (2013), identify a close relationship between comfort and trust in the context of AV acceptance, highlighting the importance of comfort for future implementations of AV services. However, the literature provides little information about this relationship. For this reason, the paper includes a review of the main factors influencing comfort in road transport in the next section, and then in the experimental study addresses the inter-relation of the two concepts.

The literature does not offer a unanimous definition of comfort (Bellem et al., 2018). However, de Looze et al. (2003) identified the most common factors as being: (1) comfort (like trust) is subjective, (2) comfort is influenced by external factors influencing the body and the body’s response to those influences (internal), and (3) comfort is experienced as a reaction to something. Comfort in private passenger cars has been important in the literature to date and has focussed on the physical parameters of thermal, acoustic, and vibrational comfort. These in turn take into account a wider set of factors including temperature, noise, humidity, lighting, driving position, and the duration of the exposure to each of these factors (Zuska and Więckowski, 2018). The most important factors that influence comfort perceived by both drivers and passengers have been identified as acceleration and vibration (Eriksson and Friberg, 2000, Lin et al., 2010), with the latter being potentially injurious for humans (Stańczyk and Zuska, 2015), especially at high speeds. Hence, speed has a negative impact on comfort when driving in a car (Uys et al., 2007, Barone et al., 2016) or riding on a bus (Bodini et al., 2014, Barone et al., 2018). The increased vibrations at higher speeds make drivers feel uncomfortable (Hu et al., 2017), and feeling less safe at high speed can lead drivers to reduce speed (Branzi et al., 2017). This indicates speed might have a negative relationship with comfort, and suggests it would also have a negative influence on trust, at least in some driving conditions. Fig. 2 shows the relationship between perceived risk and speed, which has impacts on infra sound, noise and vibrations and an indirect but significant effect on comfort.

Within the literature on private car comfort there has been a focus on the driver and the driver’s seat, emphasising the active role of driving within the vehicle, with relatively little concern for passenger comfort (Erol et al., 2014, Bellem et al., 2018), despite passengers not having the cognitive load of the driving task to ‘distract’ them, so their comfort perceptions potentially being more acute. Indeed, this difference in comfort sensitivity has been found in the previous studies which did consider passengers (Tan, 2005, Fitzpatrick et al., 2007), which suggest that car passengers usually experience higher discomfort at lower rates of acceleration than car drivers do, probably because they are involved in different tasks to drivers during the journey. Also, the degree of ‘jerk’, or lack of smooth progress, has been widely recognized as a determinant of passenger comfort (Le Vine et al., 2015, Bellem et al., 2016). Hence, it is argued that a refocus on passenger comfort is important to understand AV acceptance, particularly if considering that an uncertain but potentially large proportion of AV users who would normally have expected to take the role of driver will become passengers.

If a key relevance of the car comfort literature arises due to the aspiration that current car drivers will become SAV passengers, it is nonetheless important to consider the road public transport comfort literature, due to the potential similarities with aspects of the SAV service model, for example, the collective nature of the services, and their need for some form of stops and access management. Comfort has been recognised as an important factor influencing perceived satisfaction with public transport services (Dell’Olio et al., 2011, Fellesson and Friman, 2012, Beirão and Cabral, 2007, Lin et al., 2010). In addition to the above-mentioned factors influencing comfort on cars, comfort on buses can depend on the availability of soft and clean seats, an in-vehicle temperature range identified as pleasant, and a low occupancy factor (Beirão and Cabral, 2007), with crowding increasing perceptions of risk to personal safety and security (Cox et al., 2006, Katz and Rahman, 2010), which can increase anxiety (Cheng, 2010) and stress (Lundberg, 1976, Mohd Mahudin et al., 2011). Crowding can also cause a feeling of invasion of privacy (Wardman and Whelan, 2011) and possibly ill-health (Cox et al., 2006, Mohd Mahudin et al., 2011).

The AV passenger experience is likely to be different from the human-driven experience, so research questions emerge as to whether differences in perceived comfort will arise, and more generally, whether the difference in style will be universally welcomed. It is possible that the AV experience will be in conflict with expectations born from habituated experience, meaning that users experience loss of control (Elbanhawi et al., 2015) or perceive the driving style as insufficiently assertive. Within the SmartShuttle project carried out in Switzerland, researchers identified positive attitudes towards the use of SAVs. However, many participants affirmed that the low operational speed of the SAV did have a negative impact on other traffic (Eden et al., 2017), highlighting the high importance of speed for a successful SAV implementation. Furthermore, within their project with SAV shuttles in Germany, Nordhoff et al. (2018) found that (low) speed had a negative impact on comfort and acceptance. The relationship between speed, comfort and trust is complex, however, as other studies showed that low speed can be positively appraised because it increases perceived safety (Bekhor et al., 2003, Rodríguez, 2017). At the same time, low speed can have a negative impact on users’ satisfaction with the experience, due to the longer travel times (Bekhor et al., 2003, Krueger et al., 2016, Nordhoff et al., 2019). The relative influence of these factors will vary according to context, notably whether operating on a shared-space campus environment or on roads with faster-moving traffic, and is likely to change as SAV shuttle competences grow and speeds can increase.

A further specific debate refers to the likely incidence of motion sickness on comfort. Notably, a review published subsequently to the experimentation for the present paper (Iskander et al., 2019) also investigated the factors that can cause ‘autonomous carsickness’, finding that nausea can have a strong impact on comfort in AVs. Iskander et al. identified passenger-related factors and vehicle-related components that can be responsible for autonomous carsickness. Among the vehicle-related components, they identified change in vehicle speed and direction (horizontal orientation), together with the levels of vertical vibration and temperature and seat and viewing position, in particular whether the latter allows visual motion information from outside the vehicle to be perceived. Sivak and Schoettle (2015) observe that the frequency and severity of motion sickness could potentially decrease if self-driving vehicles do indeed provide a smoother ride than conventional vehicles. However, Diels et al., 2016, Krause et al., 2016 explain that passengers are less able to predict the ‘oncoming motion profile’ (the expected speed and acceleration/deceleration characteristics typical of a vehicle-driver combination) when travelling in a SAV, and so can feel conflicting motion cues when engaged in non-driving tasks (e.g. reading during the journey). Diels et al. (2016) suggest that motion sickness might be experienced by as much as 50–75% of the population under these conditions. According to Iskander et al. (2019), motion sickness could represent a significant issue with fully-automated road travel, as drivers cede control over the motion of their vehicles, and even drivers who never experienced motion sickness before could become susceptible, due to their reduced attention towards the vehicle’s motion and progress.

Given the importance of having a view of the oncoming road, it can be expected that being seated travelling facing backwards in a SAV would result in increased incidence of motion sickness and therefore reduced comfort. As considered further in the methodology section below, in the current study eight comfort factors were measured in the experiment, whilst direction of travel/face was a key variable, and nausea was monitored before, during and after the experiment. Subsequent to the experimental work taking place, Nordhoff et al. (2019) reported qualitative findings from shuttle riders in Berlin confirming that being seated backwards was identified as less comfortable.

The need for synchronously-shared mobility services has been identified as a policy imperative for governance of the transition to AVs. An emerging SAV implementation niche is for small-to-medium size vehicles to operate at slow speeds in environments with some regulation over interactions with other users of the space. Some initial findings from SAV demonstration studies note relationships between speed, comfort, and trust, but these are partly an artefact of the limited capabilities of the prototype services and there are some contradictions as to whether greater or lesser speed promotes trust.

The wider literature to date on comfort and trust, some of which in the context of automation more generally, has identified many factors which contribute to driver and passenger ratings, although with an emphasis on the perceptions of car drivers. Trust is strongly subjective in being influenced by individual-experiential factors, although information provision can have a positive influence as an alternative to direct experience, and current passengers were identified as more accepting of road transport automation than drivers. Both comfort and trust are reduced by relatively high-speed travel, with greater vibration being one explanatory factor. Comfort is also subjective, with many influencing factors, but key points for the present analysis are that passengers experience comfort differently from drivers, partly due to their focus of attention and cognitive load, and the difference in automated versus human driving styles will likely affect perceptions, in different ways for different perceivers. Direction of face was implicated as a specific factor influencing comfort in a previous study. Reviews of comfort factors have also predicted that motion-related nausea will be a significant problem in AVs, particularly if passengers do not attend to the oncoming motion profile.

Having identified knowledge gaps, and given the rare opportunity to undertake research with a prototype Level 4 SAV (capable of fully automated operation on a specific route) at a closed test site, an experiment was designed primarily to measure trust and comfort perceptions in the context of a social environment of unfamiliar travellers and with the manipulation of two variables identified in the literature (speed and direction of face). In addition, the incidence of nausea was appraised and the ratings of people who mainly travelled in daily life as a car driver, or in another way, were compared. The research opportunity enabled a contribution to the relatively small evidence base of experimental studies using actual vehicles. The experimental design and conduct are explained in the next section, whilst Section 3 presents the results, Section 4 the discussion, and Section 5 the conclusions and implications.

Section snippets

Methodology

Xu et al. (2018) have argued that the understanding of public acceptance of AVs is over-dependent on online surveys that involve participants which have rarely had any experience with an AV, potentially reducing the validity of the findings. Similarly, the use of simulator studies for user-related research on AVs has advantages and disadvantages in terms of technological limitations (Payre et al., 2017). Therefore, following Xu et al., (2018), for the present study the experience of an actual

Results

This section presents the results of a range of data analyses the authors performed to understand how specific attributes of the experimental exposure affected comfort and trust ratings, and how trust and comfort are related. Section 3.1 describes the approach used to carry out the analysis of the principal hypotheses, and presents the results of the effects of the independent variables ‘speed’ and ‘DoF’ on trust, comfort and nausea. Section 3.2 analyses comfort and trust variations with length

Discussion

The section presents a discussion of the results presented in the previous section, structured in the same order of presentation, considering: (1) the impact of ‘speed’ and ‘DoF’ on trust and comfort; (2) comfort and trust variations with length of exposure; and (3) participants’ expectations and valuations before and after riding in the SAV.

Conclusion

The present paper represents a contribution to the literature on AVs and SAVs based on a test-track experiment with a functioning AV, rather than a simulator-based scenario study. Hence the study contributes to the fairly small pool of experiments in which participants experience real AV, recognised as a desirable and needed methodological condition (Molnar et al., 2018).

The literature review highlighted that many factors influence comfort on and trust in AVs, but also the high importance of

CRediT authorship contribution statement

Daniela Paddeu: Conceptualization, Investigation, Formal analysis, Methodology, Writing - original draft. Graham Parkhurst: Conceptualization, Investigation, Methodology, Writing - original draft, Writing - review & editing. Ian Shergold: Conceptualization, Investigation, Methodology, Writing - original draft, Writing - review & editing.

Acknowledgements

The research reported in this paper was funded by Innovate UK (grant reference 103288) and undertaken as part of the CAPRI Project. The authors are grateful to members of the CAPRI Consortium for making the automated vehicle trial possible, and in particular Westfield Technology Group, Fusion Processing Limited, AECOM Ltd, and YTL Property Holding (UK) Ltd. The authors also thank Dr Miriam Ricci, Mr Jonathan Flower and Mr Steven Russell for assistance with running the experiment, and our

References (95)

  • C. Gold et al.

    Trust in automation–Before and after the experience of take-over scenarios in a highly automated vehicle

    Proc. Manuf.

    (2015)
  • J. Iskander et al.

    From car sickness to autonomous car sickness: a review

    Transport. Res. Part F: Traffic Psychol. Behav.

    (2019)
  • S. Khastgir et al.

    Calibrating trust through knowledge: introducing the concept of informed safety for automation in vehicles

    Transport. Res. Part C: Emerg. Technol.

    (2018)
  • M. Körber et al.

    Introduction matters: Manipulating trust in automation and reliance in automated driving

    Appl. Ergon.

    (2018)
  • R. Krueger et al.

    Preferences for shared autonomous vehicles

    Transport. Res. Part C: Emerg. Technol.

    (2016)
  • S. Le Vine et al.

    Autonomous cars: the tension between occupant experience and intersection capacity

    Transport. Res. Part C: Emerg. Technol.

    (2015)
  • J.D. Lee et al.

    Trust, self-confidence, and operators’ adaptation to automation

    Int. J. Human-Comput. Stud.

    (1994)
  • M.W. Levin et al.

    A linear program for optimal integration of shared autonomous vehicles with public transit

    Transport. Res. Part C: Emerg. Technol.

    (2019)
  • X. Liang et al.

    Optimizing the service area and trip selection of an electric automated taxi system used for the last mile of train trips

    Transport. Res. Part E: Logist. Transport. Rev.

    (2016)
  • W. Liu

    An equilibrium analysis of commuter parking in the era of autonomous vehicles

    Transport. Res. Part C: Emerg. Technol.

    (2018)
  • L.J. Molnar et al.

    Understanding trust and acceptance of automated vehicles: An exploratory simulator study of transfer of control between automated and manual driving

    Transport. Res. Part F: Traffic Psychol. Behav.

    (2018)
  • S. Narayanan et al.

    Shared autonomous vehicle services: a comprehensive review

    Transport. Res. Part C: Emerg. Technol.

    (2020)
  • S. Nordhoff et al.

    User acceptance of automated shuttles in Berlin-Schöneberg: a questionnaire study

    Transport. Res. Part F: Traffic Psychol. Behav.

    (2018)
  • S. Nordhoff et al.

    What impressions do users have after a ride in an automated shuttle? An interview study

    Transport. Res. Part F: Traffic Psychol. Behav.

    (2019)
  • I.Y. Noy et al.

    Automated driving: safety blind spots

    Saf. Sci.

    (2018)
  • Y. Shen et al.

    Integrating shared autonomous vehicle in public transportation system: a supply-side simulation of the first-mile service in Singapore

    Transport. Res. Part A: Policy Pract.

    (2018)
  • P.E. Uys et al.

    Suspension settings for optimal ride comfort of off-road vehicles travelling on roads with different roughness and speeds

    J. Terramech.

    (2007)
  • R. Vosooghi et al.

    Shared autonomous vehicle simulation and service design

    Transport. Res. Part C: Emerg. Technol.

    (2019)
  • Z. Wadud et al.

    Help or hindrance? The travel, energy and carbon impacts of highly automated vehicles

    Transport. Res. Part A: Policy Pract.

    (2016)
  • Z. Xu et al.

    What drives people to accept automated vehicles? Findings from a field experiment

    Transport. Res. Part C: Emerg. Technol.

    (2018)
  • Alessandrini, A., 2016. CityMobil2. Experiences and recommendations. Available at:...
  • A. Alessandrini et al.

    CityMobil2: challenges and opportunities of fully automated mobility

  • V. Barone et al.

    Vibrational comfort on board the vehicle: influence of speed bumps and comparison between different categories of vehicle

    Adv. Acoust. Vib.

    (2016)
  • Bekhor, S., Zvirin, Y., Tartakovsky, L., 2003. Investigating user acceptance of cybernetic cars for a university...
  • I. Bodini et al.

    Techniques for on-board vibrational passenger comfort monitoring in public transport

    ACTA IMEKO

    (2014)
  • L. Carter et al.

    The utilization of e-government services: citizen trust, innovation and acceptance factors

    Inform. Syst. J.

    (2005)
  • Y.-H. Cheng

    Exploring passenger anxiety associated with train travel

    Transportation

    (2010)
  • J.K. Choi et al.

    Investigating the importance of trust on adopting an autonomous vehicle

    Int. J. Human-Comput. Interact.

    (2015)
  • M.P. de Looze et al.

    Sitting comfort and discomfort and the relationships with objective measures

    Ergonomics

    (2003)
  • C. Diels et al.

    Motion sickness in automated vehicles: the elephant in the room. In Road Vehicle Automation 3

    (2016)
  • C. Diels et al.

    Designing for comfort in shared and automated vehicles (SAV): a conceptual framework

    Available at

    (2017)
  • X. Dong et al.

    Transit user perceptions of driverless buses

    Transportation

    (2019)
  • G. Eden et al.

    Expectation and experience: passenger acceptance of autonomous public transportation vehicles

  • M. Elbanhawi et al.

    In the passenger seat: Investigating ride comfort measures in autonomous cars. IEEE Intelligent

    Transportation.

    (2015)
  • P. Eriksson et al.

    Ride comfort optimization of a city bus

    Struct. Multidiscip. Optim.

    (2000)
  • T. Erol et al.

    Effects of appearance on the perceived comfort of automotive seats

    Adv. Ergon. .Des., Usab. Special Populat.: Part II

    (2014)
  • C. Fornell et al.

    Evaluating structural equation models with unobservable variables and measurement error

    J. Mark. Res.

    (1981)
  • Cited by (0)

    View full text