Enhancing eco-safe driving behaviour through the use of in-vehicle human-machine interface: A qualitative study

Transportation Research Part A 100 (2017) 247–263

Contents lists available at ScienceDirect

Transportation Research Part A journal homepage: www.elsevier.com/locate/tra

Enhancing eco-safe driving behaviour through the use of in-vehicle human-machine interface: A qualitative study Atiyeh Vaezipour a,b,⇑, Andry Rakotonirainy a,b, Narelle Haworth a,b, Patricia Delhomme c a

Queensland University of Technology (QUT), Centre for Accident Research Road Safety-Queensland (CARRS-Q), 130 Victoria Park Road, Queensland 4059, Australia Queensland University of Technology (QUT), Institute of Health and Biomedical Innovation (IHBI), 60 Musk Avenue, Kelvin Grove, Queensland 4059, Australia c French Institute of Science and Technology for Transportation, Development, and Networks (IFSTTAR), Department of Planning, Mobilities, and Environment, Mobility, 25 allée des Marronniers, CS 9058 Versailles, France b

a r t i c l e

i n f o

Article history: Received 23 January 2017 Received in revised form 28 April 2017 Accepted 28 April 2017

Keywords: Eco-driving Safe driving Human-machine interface Technology Acceptance Model User-centred design

a b s t r a c t Background: The widespread reliance on motor vehicles has negative effects on both the environment and human health. The development of an innovative in-vehicle humanmachine interface (HMI) has the potential to contribute to reducing traffic pollution and road trauma. Aim: A qualitative study, using a driver-centred design approach, was carried out to test how best to provide ecological and safe (eco-safe) driving advice and feedback to drivers on their driving style via an in-vehicle HMI. Method: A total of 34 drivers (52.9% males), aged 19–61 years, participated in focus groups which explored concepts from the Technology Acceptance Model (Davis, 1989). Findings: Main themes emerging from the focus groups were: (i) perceived importance of eco-safe driving behaviour; (ii) perceived usefulness of eco-safe in-vehicle HMIs; (iii) intentions to use an eco-safe in-vehicle HMI; (iv) perceptions toward eco-safe in-vehicle HMI design characteristics; and (v) potential problems associated with using eco-safe invehicle HMIs. Implications: This study provides the foundation to inform the design and development of an evidence-based in-vehicle eco-safe HMI with high levels of driver acceptance. Recommendations for future research are also discussed. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Climate change and the impact of humankind on the environment is a prominent public health issue requiring immediate action (Gore, 2006). Ambient air pollution contributes to 3.7 million deaths each year worldwide, making it the single largest threat to environmental health on a global scale (World Health Organization, 2014). Our ever-increasing reliance on motor vehicles contributes to this problem by increasing gaseous and particulate pollution, which in turn has severe impacts on human health, such as increasing lung and heart disease. Moreover, 19% of all carbon dioxide (CO2) emissions are caused by road vehicles, with such emissions intrinsically linked to rising global temperatures (Matthews et al., 2009).

⇑ Corresponding author at: Queensland University of Technology (QUT), Centre for Accident Research Road Safety-Queensland (CARRS-Q), 130 Victoria Park Road, Queensland 4059, Australia. E-mail addresses: [email protected] (A. Vaezipour), [email protected] (A. Rakotonirainy), [email protected] (N. Haworth), patricia. [email protected] (P. Delhomme). http://dx.doi.org/10.1016/j.tra.2017.04.030 0965-8564/Ó 2017 Elsevier Ltd. All rights reserved.

248

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

There have been a number of recent developments within the transportation industry aimed at, among other things, reducing the impact of motor vehicles on the environment, such as electric or automated vehicles (Fagnant and Kockelman, 2015; Wadud et al., 2016). However, improvements associated with these initiatives have been relatively incremental due to the typically higher costs to users associated with such options. On the other hand, eco-driving, which can broadly be defined as driving behaviours aimed at reducing fuel consumption and subsequent emissions, has been demonstrated as a promising and cost-effective approach (Barkenbus, 2010; Pampel et al., 2015). The general objective of the present study is to identify user requirements and perceived acceptability toward an in-vehicle HMI system for both eco-driving and safe driving behaviours. This paper is divided into four main sections, the first of which has three sub-sections. The first section focuses on ecodriving as a climate change initiative, user-centred design and driver acceptance, and the aims of the present study. The second section outlines the study methodology, while the third section presents the results and discusses the findings. The final section outlines the conclusions of the research and its real-world implications. 1.1. Eco-driving as a climate change initiative Prior research has identified three levels of decisions associated with eco-driving (Alam and McNabola, 2014): (i) strategic decisions (e.g., vehicle selection, maintenance schedules); (ii) tactical decisions (e.g., route selection, vehicle loading); and, (iii) operational decisions (e.g., driving style). While all of these decisions are important, improving driving style is associated with relatively immediate impacts on fuel consumption and emissions once the appropriate driving style is adopted (Martin et al., 2012). Despite these advantages, many drivers do not adopt or practice eco-driving. Thus, Barkenbus (2010) described ecodriving as an underutilised initiative for battling climate change, reporting that improvements in eco-driving have the potential to reduce fuel consumption by up to 10%. As a result, governments in a number of highly motorised countries have begun to develop and implement eco-driving policies within the transport sector in a bid to reduce fuel consumption and subsequent emissions (Alam and McNabola, 2014). However, a number of studies have argued that eco-driving behaviour may at times compromise safe driving. For example, maintaining a constant cruising speed and choosing the highest appropriate gear, may increase the likelihood that a driver will decrease headway to the vehicle in front and reduce their ability to brake appropriately, in turn increasing the risk of rear-end collisions (Young and Birrell, 2012; Young et al., 2011). In addition, such behaviour may also increase the likelihood of a driver manoeuvring at inappropriately high speeds, such as when cornering (cited in CIECA, 2007). These findings highlight that the development of any eco-driving initiative should be conducted with driver safety as a critical consideration. That is, the development of in-vehicle HMI systems should ultimately seek to address eco-safe driving behaviours, defined as those behaviours that reduce fuel consumption and subsequent emissions, while also considering the impact of system use on safe driving behaviours. In recent years a number of guidelines (e.g. EcoDrivingUSA; ECOWILL, 2014), public education campaigns and driver licence training (ECODRIVEN; Graves et al., 2012 for examples) have been developed to encourage eco-driving behaviour among the driving population, with mixed results. An explanation for this may be that while many drivers are ultimately aware of the range of eco-driving behaviours that impact upon fuel consumption and subsequent emissions, they lack the technical understanding of how to appropriately perform these behaviours (Delhomme et al., 2013; Pampel et al., 2015), or may make conscious decisions to drive in a manner that is not fuel efficient or safe due to a variety of reasons they believe justify the behaviour in the given moment, such as (Harvey et al., 2013) running late or enjoying the feeling of driving fast. Alternatively, the complexity of the driving task might mean that drivers are not always aware of their actions, and in turn may not use their eco-driving knowledge and skills to their full potential (Pampel et al., 2015). For this reason, the development of in-vehicle human-machine interfaces (HMI) represent a promising approach for providing relevant real-time advice and feedback to drivers to assist them to better adopt eco-driving behaviours. Previous research has highlighted the potential for eco-driving in-vehicle HMIs to have positive impacts on fuel consumption and vehicle emissions (e.g. Ecodriver, 2011; Jonkers et al., 2016; Larsson and Ericsson, 2009; Van der Voort et al., 2001). However, it is important that the development of such initiatives adopt a user-centred design approach and carefully consider the impact of issues associated with driver acceptance on subsequent system effectiveness. 1.2. User-centred design and driver acceptance The effectiveness of an in-vehicle HMI is highly dependent on driver acceptance of the technology, and in particular on the perceived usefulness and intention to use the system. Driver acceptance has been defined as ‘‘the degree to which an individual incorporates the system in his/her driving” (Adell, 2009, p. 31). This concept has also been used to describe how much drivers would use the system and their willingness to pay to purchase a system (Jamson, 2010). Adopting a user-centred design approach can increase the likelihood of driver acceptance of an in-vehicle HMI and motivate greater usage, in turn enhancing the effectiveness of a system (Maguire, 2001). The International Organisation for Standardization ISO (2010) has highlighted four key principles and recommendations for user-centred design. Overall, these principles characterise the process of user-centred design into a number of stages (see Fig. 1). Specifically, it is argued that system developers must (i) analyse and comprehensively understand the context of the

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

249

Fig. 1. The User-Centred Design Approach (adapted from ISO, 2010).

whole user experience, such as the tasks involved, the organisational, social and environmental context of use, and the needs, motivations and limitations of the intended users; (ii) actively involve users in all stages throughout the development process; (iii) conduct user-centred evaluations; (iv) understand that the process is iterative and refine systems using the same user-centred approach; and (v) adopt a multidisciplinary approach that includes a design team with a variety of skills and perspectives. Therefore, the design and development of in-vehicle HMIs must carefully consider user requirements (Vilimek and Keinath, 2014). Moreover, Norman and Draper (1986, p. 61) argue that it is important for user requirements to be considered as early in the design process as possible, stating that ‘‘from the point of view of the user, the interface is the system, the needs of the users should dominate the design of the interface, and the needs of the interface should dominate the design of the rest of the system”. Evaluating driver acceptance of in-vehicle HMIs is therefore critical. Indeed, Sauer (1993, p. 4) argued that the acceptance and subsequent adoption of a technological system is ‘‘the single, most important factor in determining success or failure of information systems and technologies”. One of the most popular models used to understand the determinants of user acceptance of technology is the Technology Acceptance Model (TAM). This approach focuses on how perceptions regarding the usefulness and ease of use of computer information systems impact upon attitudes toward use, intentions to use and subsequent actual system use (Davis, 1989). The TAM has been identified as being a highly reliable, valid and robust approach to measuring instrument design (Taylor and Todd, 1995; Venkatesh et al., 2003). 1.3. Study aims This study addresses the scarcity of current research examining driver acceptability of in-vehicle HMIs designed to enhance eco-driving. An exploratory focus group study, based on the TAM, was conducted given that focus groups represent an efficient approach to eliciting comprehensive, in-depth discussions from participants (Braun and Clarke, 2006), while also maintaining a user-centred design approach. This user-centred, qualitative method has also been argued to be an effective approach for studying scarcely researched topics and phenomena (Given, 2015). This study aims to identify best-practice principles for the design and development of in-vehicle HMI systems to improve eco-driving, while also considering the potential implications of system use on safe driving behaviours and driver acceptance of such systems. Secondly, the study aims to explore driver perceptions and attitudes toward eco-safe behaviour, driver acceptance of, and intention to use eco-safe HMI systems, as well as attitudes toward various design and feedback characteristics. 2. Method 2.1. Participants Participants were recruited from the city of Brisbane (Queensland, Australia) primarily through university email lists, snowball sampling methods, social media posts and personal approaches at university campuses, where flyers were distributed in hardcopy. This study has a research ethics committee approval in accordance with the Australian Code for the Responsible Conduct of Research (Approval number 1500000683).

250

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

A total of 34 drivers (52.9% males) agreed to participate across six focus group sessions involving between five and six participants each. The sample size was determined after data saturation, whereby limited new additional information was being obtained from each additional focus group session (Guest et al., 2006; O’Reilly and Parker, 2012; Walker, 2012). The advantages of qualitative methods for eliciting rich, in-depth data have been clearly demonstrated in studies and the golden rule for sample sizes consider this sample size is acceptable and that additional participants would have been unlikely to have added new information (Delhomme and Meyer, 2002). As can be seen in Table 1, the sample was aged between 19 and 61 years (M = 32.11, SD = 10.44 years). The majority of participants (82.4%) held an Open driver’s licence (i.e., unrestricted), while 11.8% were Provisional licence holders (i.e., unsupervised but restricted) and 5.9% held a Learner’s licence. Roughly a third of participants (29.4%) reported being very experienced with technology, while 47.1% reported having average experience and 23.5% categorized themselves as inexperienced. 2.2. Procedure and materials Participants gave written consent to participate in audio-recorded focus group discussions and were informed that their participation was entirely voluntary and anonymous. Participants initially completed a pen-and-paper survey which collected information regarding demographics, driving experience and experience with technology. Focus groups ran for approximately 60 min and participants received a movie voucher as reimbursement for their time and contribution. Each focus group session was conducted by the same facilitator, together with another researcher who took notes. The focus group sessions were audio-recorded and transcribed. The facilitator began each focus group by outlining a range of fuel-efficient and safe driving behaviours, including smooth acceleration and deceleration, fuel efficient and consistent speed choice, anticipating traffic flow, keeping a safe headway and lane-keeping. Participants were encouraged to discuss which of these behaviours they would be interested in monitoring while driving. Following this, a series of eco-safe HMI concepts were presented to participants and briefly described with the aid of images as stimuli (see Vaezipour et al., 2016 for more details). A series of open-ended questions were then asked and participants provided feedback regarding each in-vehicle interface, including what they believed to be the key effective features and areas for improvement. Questions were semi-structured and based on the TAM focus group discussion guide cited in Mitsopoulos-Rubens and Regan (2014). Questions predominantly focused on acceptability, including: (i) user perceptions of the conceptual designs (e.g., engagement, ease of understanding and perceived ease of use); (ii) perceived impacts on eco-safe driving behaviours; (iii) perceived difficulties associated with the interface and ideas for improvements; and, (iv) intentions to use and purchase the interface. 2.3. Data analysis The audio-recorded focus groups were transcribed and imported into the Nvivo 11 software program based on principles of the TAM, as well as other key themes evident in the existing literature (Miles et al., 2013). A thematic analysis was conducted to identify patterns across responses and their relation to research questions (Braun and Clarke, 2006; Mark and Yardley, 2004). The transcriptions of focus group sessions were coded using a thematic content analysis (Weber, 1990), in order to reduce the information to more relevant, manageable pieces of data. All ideas, expressions, and words were classified into content themes on the coding grid. Each defined theme consisted of units comprising one or more words with similar meaning (Weber, 1990). Emerging themes that did not fit within this framework were also considered during the coding process. This review and refinement process led to the development of the final themes and sub themes.

Table 1 Demographics of focus group participants. Variable

No. of participants

Variable

No. of participants

Gender Female Male

16 18

(47.1%) (52.9%)

Age 18–29 30–39 40 +

16 13 5

(47.1%) (38.2%) (14.7%)

Licence type Learner Provisional Open

2 4 28

(5.9%) (11.8%) (82.4%)

Motivation to improve fuel efficiency Environment & money Save money Environment Nothing

24 8 1 1

(70.6%) (23.5%) (2.9%) (2.9%)

Average hours driven < 5 h/week 6–10 h/week 11 +

12 10 12

(35.3%) (29.4%) (35.3%)

Experience with technology Experienced Average Inexperienced

10 16 8

(29.4%) (47.1%) (23.5%)

251

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

Another researcher also coded the transcriptions in order to validate the initial interpretations. Discrepancies were thoroughly discussed and guided the thematic development, so as to generate a set of coherent themes that reflected a comprehensive and precise set of meanings for the participants’ comments. Improvements to the themes were made until the intercoder agreement was 100%. 3. Findings and discussion 3.1. Number of statements, themes and sub-themes The sample of 34 participants produced 273 statements, at a mean rate of 8.03 statements per participant. The statements were defined as participant attitudes and perceptions in relation to each theme or sub theme. A total of five themes emerged from the focus group discussions: (1) the perceived importance of monitoring eco-safe driving behaviours; (2) perceived usefulness of eco-safe in-vehicle HMIs; (3) intentions to use an eco-safe in-vehicle HMI; (4) perceptions toward eco-safe in-vehicle HMI design and feedback characteristics, and (5) perceptions toward potential problems associated with using eco-safe in-vehicle HMIs. Thirteen sub themes were also identified (see Table 2). A chi-square test of goodness of fit was performed to determine whether there were significant differences in rates of responding by gender or age. Males cited significantly more statements than females (v2 = 10.4, p < 0.01), however no significant differences in response rates were observed by age (v2 = 4.42, p = 0.11). Tables 3–7 outline key statements from participants in relation to each of the themes and sub-themes. Direct quotes are labelled with the corresponding age, gender (i.e. F or M), and the number of years each participant had held their driver’s licence (e.g., 34, F, 14) refers to a 34 year old female who obtained her driver’s licence 14 years ago. 3.2. Perceived importance of monitoring eco-safe driving behaviours Participants were asked to discuss which behaviours they perceived as important for improving fuel consumption and safe driving, specifically in relation to those behaviours that could be monitored by an in-vehicle eco-safe HMI system. Ten participants made a total of 12 statements regarding the perceived importance of monitoring eco-safe driving behaviours. Table 3 outlines the key relevant quotes related to the perceived importance of eco-safe driving behaviours. Overall, participants generally supported the notion that being able to monitor eco-driving and safe driving behaviour is important and that an eco-safe in-vehicle HMI system has the potential to improve their own driving by increasing their awareness of their behaviour. Monitoring was argued to be particularly beneficial in relation to behaviours that are typically difficult for one to subjectively monitor themselves, such as acceleration and deceleration (Table 3, a.i), as well as behaviours perceived to have a considerable impact on safety, such as maintaining a safe headway, lane keeping and speeding (Table 3, b.i). Taken together, these findings add support to the hypothesis that a well-designed, user-centric system would be likely to be associated with relatively high levels of acceptance among drivers.

Table 2 Summary of thematic analysis. Key Themes and sub-themes

Number of participants

Frequency of statements

Age 18–29

30–39

40 +

Gender M

F

1. Perceived importance of eco-safe driving behaviours

10 (6M, 4F)

12 (4.3%)

6

3

3

8

4

2. Perceived usefulness of system a. Save fuel and money b. Improve safe driving c. Help environment

20 (11M, 9F) 12 (6M, 6F) 10 (4M, 6F) 3 (2M, 1F)

30 (10.9%) 16 (5.8%) 14 (5.1%) 3 (16.1%)

9 5 9 2

15 7 2 1

6 4 3 0

18 10 7 2

12 6 7 1

3. Intentions to use system a. Driver type b. Perceived ease of use/usability c. Accessibility and cost d. Normative influences

22 (13M, 9F) 6 (4M, 2F) 15 (9M, 6F) 4 (2M, 2F)

44 (16.1%) 7 (2.5%) 19 (6.9%) 5 (1.8%)

21 5 14 4

13 1 5 1

10 1 0 0

26 4 12 3

18 3 7 2

4. Design characteristics a. Device characteristics b. Feedback characteristics

18 (10M, 8F) 20 (12M, 8F)

36 (13.1%) 34 (12.4%)

20 12

15 12

1 10

24 23

12 11

5. Potential problems a. Distraction b. Privacy concern c. Perceived accuracy and trust d. Novelty e. Promote unsafe/illegal behaviour

11 (6M, 5F) 12 (9M, 3F) 7 (4M, 3F) 7 (2M, 5F) 5 (3M, 2F)

15 (5.4%) 15 (5.4%) 7 (2.5%) 9 (3.2%) 7(2.5%)

2 5 2 0 4

2 1 1 2 3

9 12 4 4 5

6 3 3 5 2

11 9 4 7 0

252

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

Table 3 Selected quotes relevant to theme 1 - ‘‘Perceived importance of eco-safe driving behaviours”. Subtheme a. Fuel consumption

b. Safety

c. Optimism bias

d. Monitoring other drivers

Relevant Quotes i. ‘‘I think monitoring the acceleration and deceleration would be effective because I don’t know myself what’s the right stopping distance and acceleration distance. I know it from a point of view of a passenger who drives with me, just from feedback, they believe that I stop and start quite quickly. So I think monitoring that will be quite effective for my own personal benefit” (22, F, 1). i. ‘‘I think keeping a safe headway, that’s important, especially for safety. Not that I have an issue with that myself, but I do have a couple of friends who do keep quite an unsafe distance between the cars and they don’t realise themselves, so I think that would be effective” (22, F, 1). ‘‘I like having an awareness of where I am in a lane, maybe some information about vehicles behind me and blind spots will be really useful. I like the idea of where I am in a lane, behind me, in front of me, things I can’t always see” (32, M, 15). i. ‘‘I think most people would think that they are a safe driver. If you ask people to rate their own driving, I’m pretty sure they’ll think they’re safe. Everyone thinks everyone else is a bad driver” (39, M, 21). ‘‘I would say safety for everyone else because that’s the first thing I learned as a driver: don’t trust anyone on the road – they can do the wrong thing. But for myself, I trust myself, it’s just whether or not I can drive more fuel efficiently” (29, M, 10). i. ‘‘I think if it showed like they’re [other drivers] moving out of lanes, or they’re speeding or they’re braking suddenly, that kind of thing, as far as safety could be. . .more helpful to say that car ahead of you is doing something that you might need to be a little bit more aware of” (33, F, 13).

Table 4 Selected quotes relevant to theme 2 - ‘‘Perceived usefulness of an eco-safe in-vehicle HMI”. Subtheme

Relevant Quotes

a. Save fuel and money

i. ‘‘I’m interested in the fuel efficiency stuff because I don’t have anything like that in my car but if I did, that would be useful” (32, M, 15). ‘‘Save money. Basically, it’s the efficiency and eco-driving. Petrol is expensive so if you do enough driving it [the invehicle system] will save you quite a bit of money (61, M, 40). ‘‘With sudden stops and starts it’s often more impact on brakes and more, sort of, wear and tear on the mechanics of the car which there’s a cost associated with, maintenance costs” (35, F, 17).‘‘My car doesn’t have any technology at all. So I’ll be interested in knowing how much petrol am I using or when I went on a long drive, how much did I waste, so I’d be interested in it just for the feedback” (39, M, 21). ii. ‘‘I had this argument with one of my friends at work. He was saying ‘‘I’m a very good driver, I’ve never had an accident in my life. But in terms of being eco-friendly, he was terrible. He was absolutely terrible. I think these criteria would be good to get a sense of whether you’re driving in an eco-friendly way” (34, M, 14). iii. ‘‘If safe driving is more linked to maintenance costs and things like that, I think I would care more about it” (32, M, 15). ‘‘Some people wouldn’t care about the environment but they might care about how much money they saved” (39, M, 21).

b. Improving safe driving

i. ‘‘I think it would be hard not to improve your driving if it’s giving you visual cues to say, you’re stopping very aggressively and you go, am I really? I’ll try to stop less aggressive next time” (27, M, 9). ‘‘Could also be things that for years you thought was the right thing to do, and you’ve been doing it but maybe it’s wrong. Maybe a system like this will show you - well actually this is how you safely decelerate” (39, M, 21). ii. ‘‘I think I’d personally improve my driving because it’s quite rare that I drive with somebody else in my car, so I don’t have the opportunity to gain feedback on my own driving. And I guess that would be the same for so many other people as well who are solo drivers and don’t also have passengers. This is a good system to provide that feedback that they lack” (22, F, 1). iii. ‘‘I’m also very interested in the safety. I want to know if I’m keeping to the right speed, because obviously with children in the backseat, you want to make sure . . . My motivation would be more about safe driving because I’m carrying two children” (35, F, 17). iv. ‘‘I think it [monitoring safe driving] would be really good but I don’t think many people would like to be told that you’re not driving safely or feel that they’re being nagged about it. I think it needs to be like a subconscious thing” (61, M, 40). v. ‘‘If you’ve driven safely and you haven’t had a crash, why are you going to purchase a product? Or if you’ve never had a speeding fine or a ticket or anything, why are you going to purchase a system that’s going to tell you to drive safe when technically you’re driving safe already? (27, M, 9). ‘‘I don’t think bad drivers would give a damn” (23, F, 7). i. ‘‘I like the idea of finding out the economic and environmental benefit of your behaviour” (53, F, 35). ‘‘Carbon footprint. Yeah, tracing how you are doing in terms of your carbon footprint” (33, M, 15).

c. Help the environment

Interestingly, overall responses suggested some evidence of illusory superiority effect (McKenna and Myers, 1997; Waylen et al., 2004). Specifically, participants typically argued that, in relation to other drivers, it was most important to monitor safety behaviours, while in relation to their own behaviour, it was most important to monitor fuel efficiency (Table 3, c.i). Drivers may perceive themselves to be safer, better drivers than the average driver. While there is some evidence to suggest that constructive feedback from an in-vehicle HMI that challenges such biases may motivate behaviour change (Horrey et al., 2012), there is also the potential that such feedback may lead to cognitive dissonance (Festinger, 1962) and result in users disengaging with the system, particularly amongst those drivers who require behaviour change the most. Following on from this bias, two participants also stated that they would appreciate a system that could monitor the behaviours of other drivers, perhaps in an intuitive manner, and warn them of potentially unsafe behaviour of other drivers (Table 3, d.i). Thus existing technologies, such as hazard detection and warning systems may be a useful inclusion in the design of eco-safe in-vehicle HMI systems.

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

253

Table 5 Selected quotes relevant to theme 3 - ‘‘Intentions to use an eco-safe in-vehicle HMI”. Subtheme a. Driver type

b. Perceived ease of use and usability

Relevant Quotes i. ‘‘I do agree that it could be used to educate people, like learners . . . if you had something like this, it would be like, you’re being watched by big brother, and you cannot speed. But also that’s sort of the highest risk age group as well, 17 years to 2500 (35, F, 17).‘‘It is good because if you learn to drive efficiently when you first start driving, then you would probably do it for the rest of the time you drive, which is very good.” (28, M, 10). ii. ‘‘It will really work with families if they’ve got to share a car. Like if you are a parent and you share your car with just say, two teenagers, you’d be like, ‘okay, you can drive the car but you need to use the system. So, when you get home, I can see how you’re driving’. It would be a perfect system to go, ‘how did my son drive’. Like did he just drive it super aggressively and he’s not going to get away with that” (27, M, 9). ‘‘Parents putting it [the eco-safe in-vehicle system] into watch what the kids doing driving their car. . .I think parents may find it important for their children learning to drive” (22, F, 5). iii. ‘‘I think the people who would benefit most from this are people who think they’re good drivers but don’t have any feedback . . . but then they are still going to go, the system is wrong, I’m good . . . I don’t think it would improve my driving because I think I’d be interested to know that I wasn’t leaving enough space, but I think because I’ve been driving for 14–15 years and not rear-ended anyone. So, I’m pretty happy that I can drive at that level, whereas the system might not think I can, which is fair enough but, I don’t think I’d pay it much attention” (30, M, 13).‘‘I’d probably say no. It’s beneficial but I don’t care. I will make right choices . . . I know I’m doing what I’m doing. I know its fuel efficient, I don’t need to have a thing [in-vehicle system] tell me” (29, M, 10). iv. ‘‘I’m not into technology . . . so I probably wouldn’t [use an eco-safe in-vehicle system]. Plus, I’ve been driving for a long time . . . I think it’s brilliant for people who are comfortable with all this new technology but for me, I’m just from that era . . . I know a lot of people that are of my age that are very into new technology but I’m one of those people that can’t” (56, F, 36). v. ‘‘I’ll probably suggest it to my boss to give out to all the different drivers and stuff just because sometimes you have trades people getting back real quick from a delivery or something, but he doesn’t know what they’ve been doing – flying and just throwing the vehicle around the corners and stuff. So in that sense, I think it’s definitely got a lot of applications in businesses” (22, M, 4). vi. ‘‘Maybe use it as punishment for people, if someone’s continuously a bad driver or been drunk or reckless driving, throw it in their car and then the report is there for what they’re doing and then police can see it. So a mandatory system for some people” (29, M, 10). ‘‘I think the most valuable application for it [an eco-safe in-vehicle system] is mandatory use for bad drivers. The people that need to monitor that sort of behaviour . . . safety-wise, people that would most benefit from it, I think they’re doing the wrong thing now because they don’t care” (56, F, 36). ‘‘People with a lot of traffic infringements, lots of speeding fines, who have low points, that kind of stuff” (22, M, 4). i. ‘‘Depends how user friendly it is and how quickly I can do it [figure out how to use it]. If it’s going to be something that’s really time-consuming, maybe not [I wouldn’t use it]” (23, M, 1). ‘‘I would use it, only if it’s easy to use. Like I use cruise control all the time because it’s really simple” (23, F, 7).

c. Accessibility and cost

i. ‘‘I would use the system if I had access to it” (19, M, 1). ii. ‘‘It depends on how much it would cost. If it was free, yeah. If it cost money, probably not” (27, M, 9). ‘‘If it was supplied for free [I would use it]. I wouldn’t pay for it as an extra in my car” (29, M, 10). ‘‘I would consider it value adding for the car, like if I was looking at purchasing a brand new car, and it had it already installed and ready to go, I’d pay a little bit [for the system]” (23, F, 7). iii. ‘‘If it’s going to cost me $50 or $100, that’s a tank of fuel already. Like, unless it’s going to save me $5 a tank or $10 a tank, I’m not going to do it” (27, M, 9). ‘‘If it had studies that said it was going to save me like a $100, then I would probably pay like $10 or $20 to get it, but I’m not going to pay more than it’s worth to have” (22, F, 5). iv. ‘‘I’d pay for it, if there were rewards, I’d pay for it and then maybe drive better to gain more rewards” (30, M, 13).

d. Normative influences

i. ‘‘If I did have friends using it that would make me want to use it” (23, M, 1). ‘‘I’d probably have a try if a friend had tried it and said it’s good or if I saw it on a TV show or a television ad” (31, F, 6). ii. ‘‘If it came out in your insurance letters or when you renewed your licence, okay, check out this” (23, F, 7).

3.3. Perceived usefulness of an eco-safe in-vehicle system Perceived usefulness can be defined as the degree to which an individual believes that using a particular system (e.g., an eco-safe in-vehicle HMI) has the potential to enhance their performance on a particular task (e.g., eco-safe driving). That is, it relates to whether the system satisfies the needs and requirements of the user. Overall, participants identified three broad ways they believed an eco-safe in-vehicle HMI could be useful and help to improve their driving. This included: (a) helping them to save fuel and money; (b) improving awareness and increasing safety; and, (c) supporting eco-driving and helping the environment.

3.3.1. Save fuel and money Twelve participants made a total of 16 statements regarding the perceived usefulness of eco-safe in-vehicle HMI to save fuel and money, with the vast majority arguing that such systems could increase their ability to monitor their fuel efficiency and improve their eco-driving style. Ultimately, this was most often driven by the desire to save money through reduced fuel and vehicle maintenance costs (Table 4, a.i). A number of participants also suggested that such a system may be beneficial for drivers who perceived themselves to be ‘‘good” drivers based on their crash or infringement histories, but who had little insight into the eco-friendly nature of their driving behaviour (Table 4 a.ii).

254

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

Table 6 Selected quotes relevant to theme 4 - ‘‘Eco-safe in-vehicle HMI design characteristics”. Subtheme

Relevant Quotes

a. Device characteristics

i. ‘‘I think having it in-built [into the vehicle] will be much more effective because who’s going to opt to go and actually purchase this system?” (22, F, 1). ‘‘I think it should already be in the car when I’m getting the car. That might make me use it” (20, F, 3). ii. ‘‘See my trouble is I drive a lot of different vehicles every day. Like it’s [the eco-safe-in-vehicle system] going to have to be transferable from vehicle to vehicle” (30, M, 13). ‘‘I probably do think that having a little additional system would be the easiest way to roll it out” (22, F, 1). iii. ‘‘If it would look good in your car. Because a lot of people, like myself, I wouldn’t want something that looks disgusting on my dash even if it is going to provide safe driving . . . I don’t need to have something hanging out of my dash” (61, M, 40). iv. ‘‘Well the most efficient way would be the whole thing would be phone linked. Like that would be the easiest way to bring it out” (30, M, 13). ‘‘I think a phone is obviously the easiest way despite all its possible downsides. The majority of people who drive have a phone . . . a two minute download and you’ve got it” (22, M, 4). v. ‘‘I’ve a little bit of an issue with the whole thing being at all phone linked. Like if you’re trying to disengage phones from driving. . . in your mind you are linking a phone and a car as one thing” (30, M, 13). ‘‘I definitely think scrapping the phone idea, I don’t think it should have any relation to mobile phones at all” (22, F, 1). vi. ‘‘For the system to function, it’s going to need to have a lot of sensors to know what it’s doing. Like an iPhone’s not going to be able to tell you when you’re accelerating or decelerating” (30, M, 13). vii. ‘‘I’ve got kids fighting in the back seat and I’ve just got to get to ballet. I’m just not going to stop and do this and tap on my phone. But if it were automatically starting up in the car or somehow connected to the car, it would be okay. . .. it shouldn’t be a process of having a setup stage as you get into the vehicle. It should just be, switch it on when the car turns on” (35, F, 17). ‘‘It would need to be quick to setup because if I was in any kind of hurry, and I jump in the car, and it was going to take a minute to log on and do stuff to set it up, I’d be less likely to use it. But if it was pretty quick. . . yeah that would be quite good” (39, M, 21). viii. ‘‘You can combine it with GPS and all those kinds of things” (23, F, 7). ix. ‘‘A lot of cars now have a little screen that will show you everything and if it just popped up on that, for me, it’s just looking, it’s not on the windscreen, I’m just glancing to the side and that would be fine” (35, F, 17).

b. Feedback characteristics

i. ‘‘I think even having the options for each individual driver, to just like whether they want to customise, semi-customise because everybody’s going to have different opinions and different preferences . . . That just makes it a little bit more enjoyable, a little bit more relatable” (22, F, 1). ‘‘If you can have the audio or the visual [feedback], or both of them, some people will like the visual without the audio, or some will like the audio without visual. It needs to be personal” (61, M, 40). ii. ‘‘Both auditory and visual. Because just in case if you miss the visual, the audio will help you” (19, M, 1). iii. ‘‘I reckon visual is very quick, easier when you’re driving . . . I think that’s very quick and easy to communicate visually” (34, M, 14). ‘‘I’d prefer more audio stuff . . . I prefer to be told with a beep or something like that I’m too close to something or going outside my lane or something like that”(32, M, 15). iv. ‘‘I would like to have colours . . . so you don’t even need to look at it to know what’s going on, you just sort of see the colour and you know” (28, M, 10). ‘‘If you just keep a particular colour of light say red and green. Like if you do bad red pops up and you’re not actually looking at the colour but you can still see” (26, M, 3). v. ‘‘I keep on thinking about the night, about the lights [being distractive inside of the vehicle] in the night time” (25, M, 6). ‘‘You don’t want to punish someone if you want them to keep using your thing, so red is an aggressive colour” (23, M, 1).

The desire to use an eco-safe in-vehicle HMI to save fuel and money was noted even among participants who expressed less concern with the potential for the system to positively impact upon driving safety and/or the environment (Table 4, a. iii). Personal monetary gains from using the system may be a particularly useful aspect of the system to focus on when attempting to motivate more problematic and unsafe drivers to use the system. 3.3.2. Improving safe driving Ten participants made a total of 14 statements regarding the perceived usefulness of eco-safe in-vehicle HMI to improve safe driving. A number of participants believed such a system could enhance safety by increasing their awareness of their driving behaviour, through an increased ability to monitor their behaviour, thus alerting them to opportunities to improve their driving ability and overall safety (Table 4, b.i). This was particularly pertinent amongst drivers who typically travelled alone and thus had fewer opportunities to receive other forms of feedback regarding their driving behaviour and for drivers without technology in their vehicle to help them ascertain their fuel efficiency (Table 4, b.ii). A number of participants also suggested that the presence of passengers, and in particular children, would increase their motivation to use such an in-vehicle system to improve their driving safety (Table 4, b.iii). However, some participants suggested that direct feedback might not be well received by all drivers, given that it may conflict with their own perceptions of their driving ability and levels of safety (Table 4, b.iv). A number of participants noted that such a system might help drivers understand that there is more to being a ‘‘good” driver than a clean crash or offence record, by highlighting the efficiency, or lack thereof, of their driving (Table 4, b.v). However, they also questioned whether such drivers would be interested in using such a system. 3.3.3. Help the environment Overall, only three participants reported that they believed an eco-safe in-vehicle HMI would be useful for helping the environment. Indeed, only three statements from as many participants were made in relation to this subtheme (Table 4,

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

255

Table 7 Selected quotes relevant to theme 5 - ‘‘Potential problems associated with the system”. Subtheme

Relevant Quotes

a. Distraction

i. ‘‘Information that’s coming up on the screen, that’s a distraction. That’s no different to texting while you drive or any other distraction like that. To me it defeats the purpose of safety for something that’s trivial” (39, M, 21). ‘‘I think that [an in-vehicle system] could be distracting . . . like I know for some people it’s not distracting, but for other people, they can’t even use a GPS and look away from the road for two seconds” (22, F, 5). ii. ‘‘It can be irritating . . . I might turn it off” (51, F, 35).

b. Privacy concern

i. ‘‘I’d never use it if it any of the information was sent to other people. Never . . . If it ever goes to the police or anything like that, if you’re really being unsafe, I just wouldn’t use it” (23, M, 1). ‘‘I think also it would need to have no sort of tangible memory in regards to what you’ve been doing . . . If it [the eco-safe in-vehicle system] possibly became incriminating for me, then I won’t be using it” (30, M, 13).

c. Perceived accuracy and trust in the system

i. ‘‘It [the feedback] has to be realistic, that’s what I always say . . . Is it going to be, ‘you have to slow down in a kilometre’ when you can easily slowdown in 200 m and doesn’t make it like you’re driving aggressively. I’m not going to drive like an 80 year old man, because you are going to lose time in your day . . . If the system’s going to tell you that that’s a safe distance, you’re going to be sitting in traffic and people are going to be around you going, ‘what the hell is this guy doing? There’s like 6 car lengths in front of him that anyone could be sitting in’ . . . If it keeps coming up when you feel like, no, I’m not travelling too close, no I’m not decelerating too fast, if it’s not set to a standard that’s not insanely slow driving [then I’d be more likely to use the system]” (27, M, 9).‘‘I’d get really pissed if I got penalised because I went down beside a car and it said that it was too close. I’m like, well, if it’s not scraping, it’s not close, it fits” (29, M, 10). i. ‘‘I think in the short term I’d use it. But long-term – it depends upon the finer elements such as alerts and other little bits . . . as to whether I use it long term or not” (22, F, 1). ‘‘I don’t know for how long [I would use it]. Like it will be interesting for a little bit, but I think you lose interest, especially if you’ve fixed your bad habits” (27, M, 9). ‘‘I think for me, the problem will be the consistency of usage, like how long are you going to use it for? There’s always this period when people start using [technology] and they are super excited and after a point they are like, ‘oh, it’s too much of a hassle’” (22, M, 5). ii. ‘‘Probably once I get a sense of how I was doing, kind of like using a pedometer, it’s like ‘okay, I walk this many average steps, I don’t need to wear it anymore’. So then you go, ‘okay, I actually drive okay . . . I really don’t need that feedback anymore’” (53, F, 35).

d. Novelty

e. Promotion of unsafe and illegal driving behaviours

i. ‘‘I might try and see how I could push the system, like make it flash a lot that I’m doing things wrong and people would do that, like when you used to have the alcohol breath testers in the pub, you’d see who can get the highest reading, it would be similar thing with this” (30, M, 13).‘‘Good idea, if it doesn’t get abused or misused. I can see my teenage son . . . use it for bad behaviour, instead of improving, trying to compete in bad behaviour” (51, F, 35).

c.i). This finding is somewhat surprising given the media attention given to issues such as climate change, fossil fuels and environmental sustainability. While further research is required to more comprehensively investigate this issue, it may be that many drivers are either uninformed or overwhelmed by the myriad of information in the media, or have low perceptions regarding their ability to make a significant difference. Alternatively, drivers may hold paradoxical attitudes whereby they believe that it is important to protect the environment while also exhibiting behaviours that are not congruent with this belief. Such paradoxical attitudes have been demonstrated previously in the area of road safety, such as when drivers acknowledge that speeding is dangerous while also reporting frequent speeding behaviour (Fleiter and Watson, 2006). Following initial questions which discussed eco-safe in-vehicle HMIs in a general sense, from which the preceding themes emerged, participants were presented with a number of specific concept designs. The remaining themes outline participant perceptions and attitudes toward these concepts in particular. 3.4. Intentions to use an eco-safe in-vehicle HMI A number of factors were found to influence intentions to use an eco-safe in-vehicle HMI system. These included: a) driver type; b) perceived ease of use and usability; c) accessibility and cost; d) normative influences; and e) attitudes toward safe and eco-friendly driving. These factors are discussed in more detail in the sections below. 3.4.1. Driver type Twenty-two participants made a total of 44 statements regarding the influence of driver type on intentions to use an ecosafe in-vehicle HMI. Overall, participants agreed that particular groups of drivers would be more likely to benefit from the use of such a system compared to others. Indeed, a number of participants argued that novice drivers could use the HMI to educate themselves from the early stages of driving, improving their fuel efficiency and safety and increasing the likelihood that an eco-safe driving style would be retained (Table 5, a.i). In addition, a number of participants reported that parents of teenagers could use the system to monitor the driving behaviour of their children (Table 5, a.ii). Such an approach would serve a number of benefits, most notably motivating a key population of drivers who typically are more resistant to voluntary use of such systems and contributing to wide-scale system adoption among a new generation of drivers. This also represents an opportunity for the use of gamification principles in the design of the in-vehicle interface (Orfila et al., 2016; Steinberger et al., 2017), which could facilitate social persuasion among

256

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

family and friends, such as motivating behaviour change through competition and additional incentives (e.g. if children produce better eco-safe driving scores parents agree to pay for fuel). Moreover, parents who install such a system for their children may also be more likely to use the system themselves when driving, further increasing the benefit of the system. On the contrary, a number of participants expressed concerns that more experienced drivers may be reluctant to accept feedback from an eco-safe in-vehicle HMI system, particularly if it contradicts their own perceptions of their eco-safe driving performance. Indeed, some participants suggested that they would ignore feedback from the system if such a contradiction was observed (Table 5, a.iii). Such statements may reflect a distrust in technology to accurately evaluate driving behaviour with consideration of contextual and situational factors, or even a subconscious unwillingness to receive objective feedback associated with their optimism biases. Indeed, one participant mentioned her extensive previous driving experience and lack of skill with technology as the primary reason she would be uninterested in using such a system (Table 5, a.iv). Overall, this finding and others related to apparent optimism biases among some drivers, suggests a need to better educate drivers regarding relative versus absolute risks. That is, it appears that many drivers believe that an absence of traffic crashes or infringements from their driving history is evidence of their ability and safety as a driver. While in some cases this is likely to be a true reflection, in others it may simply be a function of the low absolute risk of involvement in a traffic crash in Australia (Bureau of Infrastructure Transport and Regional Economics, 2015), with consistent evidence highlighting the relative risks associated with a variety of illegal and unsafe driving behaviours (Peden et al., 2004). Another participant suggested that eco-safe in-vehicle HMIs may also be beneficial for use among business vehicle fleets, in particular those industries where driving between locations in a timely manner (e.g., delivery drivers) is encouraged (Table 5, a.v). Given previous research demonstrating the increased risk of vehicle crashes and traffic infringements among occupational drivers (Robb et al., 2008), the acceptability and effectiveness of such a system in vehicle fleets is worthy of further research. Indeed, such an application may have subsequent benefits if occupational drivers realise the positive impacts of the system and become motivated to use the system in their personal vehicles. It is at this point worth noting the snowball effect such a system would have in shared vehicles if all drivers were motivated to use the system, which might be likely if significant others (e.g., family members) are using the system and share positive experiences associated with its use. Finally, a number of participants argued that an eco-safe in-vehicle HMI system would be a beneficial sanction for offending drivers. That is, participants suggested making it mandatory for offending drivers to use the system in order to continue driving or following a licence suspension (Table 5, a.vi). Such initiatives are already evident in the area of road safety, such as the use of alcohol ignition interlocks (Elder et al., 2011) and suggestions for the use of mandatory intelligent speed adaptation (ISA) as a sanction for speeding offenders (Van der Pas et al., 2014). 3.4.2. Perceived ease of use and usability Six participants made a total of seven statements regarding the perceived ease of use and usability of the concept eco-safe in-vehicle HMIs on intentions to use. The degree to which participants perceived the systems as being easy to use, while also being effective, efficient and engaging, was reported as having a positive relationship with self-reported intentions to use such a system (Table 5, b.i). This finding is consistent with previous research using the Technology Acceptance Model (TAM) to predict intentions to use in-vehicle technology (Ghazizadeh and Lee, 2014). However, given the demanding nature of the task of driving, the challenge remains of how to design a system such that it meets these user requirements without distracting the user from the primary task (i.e., driving) and producing cognitive overload. Thus, the goal should be to design a system that maintains functionality (e.g., effectiveness, efficiency, user-satisfaction) while requiring minimal effort from the driver. 3.4.3. Cost and accessibility Fifteen participants made a total of 19 statements regarding the influence of cost and accessibility of the concept eco-safe in-vehicle HMIs on intentions to use. Participants reported a greater intention to use such a system if it was readily accessible to them (Table 5, c.i). Moreover, the cost of the system appeared to heavily impact upon perceptions of accessibility, with the majority of participants reporting that reduced system purchase costs would be associated with greater intentions to use the system (Table 5, c.ii). This finding is consistent with previous research (Stevens and Burnett, 2014) arguing that cost, among others factors, is a key criteria of practical acceptability. Overall, participants were quite polarised regarding their willingness to pay for such a system. While some argued that they would only be willing to use the system if it was free, others reported a willingness to purchase the system, and perceived it as a device that would add value to their vehicle (Table 5, c.ii). This finding has important implications for the design of such a system. That is, the design of the system as a smartphone application would facilitate low consumer purchase costs, but would be associated with a number of additional issues related to the device characteristics (see Section 3.4), such as automaticity of use. On the other hand, a stand-alone after-market or built-in device would allow for the design of a system that addresses best practice device characteristics, but would undoubtedly be more expensive and thus unattractive to some potential users. Encouragingly, a number of participants argued that they would be willing to invest in an eco-safe in-vehicle HMI if they were confident that the financial benefits associated with using the system, such as saving money on fuel and vehicle maintenance, would outweigh the costs of purchasing the system (Table 5, c.iii). Financial incentives, such as rewards and prizes associated with eco-safe driving behaviour, were also reported as being positively associated with intentions to use the

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

257

system (Table 5, c.iv). This finding is consistent with previous research that found a positive impact of a reward strategy on fuel efficiency among bus fleets (Lai, 2015). Such rewards could include discounts on vehicle insurance and/or registration, or prizes associated with eco-safe driving behaviour, such as movie tickets, fuel vouchers or other incentives. Effectiveness of an incentives approach has been evidenced in a trial conducted with Provisional drivers in Newcastle, Australia, as part of the Samsung S-Drive smartphone application initiative, whereby young drivers earned rewards for safe driving (Bennett, 2014). So, for incentives to be effective, the rewards need to be meaningful to the user, both in terms of the degree to which the user wants the reward and perceives the reward as being worth the time and effort required to earn it. Further research should also investigate the long-term feasibility of such reward schemes. 3.4.4. Normative influence Four participants made a total of five statements regarding normative influences on intentions to use the concept eco-safe in-vehicle HMIs. Participants highlighted the influence of their peers and social network on the likelihood that they would use the system, suggesting a greater intention to use the system themselves, if their peers used and recommended it (Table 5, d.i). Insurance companies and transport authorities were also reported as being able to influence intentions to use an HMI, with a number of participants suggesting that authoritative persuasion or recommendation could be an effective approach to encouraging system use (Table 5, d.ii). Taken together, these findings indicate the importance of a trustworthy source from which drivers can obtain evaluative information regarding the utility of the system. These findings are also consistent with previous research on the positive effect of normative influences on technology acceptance. Furthermore, following on from the previous section, the findings suggest a potential partnership with insurance companies and transport authorities to offer discounts on insurance and/or vehicle registration associated with system use and evidence of appropriate eco-safe driving styles. 3.5. Eco-safe in-vehicle HMI design characteristics Participants were also asked to discuss the specific design characteristics that might influence their likelihood of using an eco-safe in-vehicle HMI. This included a) various device characteristics, such as device type and automaticity, as well as b) feedback characteristics, such as feedback type (e.g., auditory, visual), accuracy and clarity of feedback and feedback customisation. 3.5.1. Device characteristics Eighteen participants made a total of 36 statements regarding their perceptions toward the device characteristics of the concept eco-safe in-vehicle HMIs. There was considerable variety in participant views regarding the most effective approach for presenting feedback to drivers with regards to device type. Overall, the majority of drivers reported a preference for one of three approaches: (i) an in-built system that comes installed in a new vehicle or can be retrofitted to older vehicles; (ii) a stand-alone, mobile system that can be fitted in vehicles (e.g., similar to many dash cams and GPS systems); or (iii) a smartphone application that can be linked to the vehicle. Those participants who argued that the most effective approach would be for the system to be in-built in new vehicles, with an option for retrofitting in older vehicles, appeared to be primarily motivated by ease of accessibility. Indeed, these participants generally suggested they would be unwilling to purchase such a system if it was not built into their vehicle (Table 6, a.i). However, it was noted that this approach would not be advantageous for individuals who drive more than one vehicle (e.g., both personal and work vehicles). Instead, three participants suggested that the system should ideally be mobile and transferable from one vehicle to another, such as a stand-alone mobile device, similar to many existing dash-cams and GPS systems (Table 6, a.ii). However a number of participants suggested that such an approach would need to consider aesthetics and not detract from the personality and overall look of their vehicle (Table 6, a.iii). The majority of participants acknowledged that a smartphone application approach would be the most readily accessible approach for a wider audience given the proliferation of smartphones, as well as being transferable from vehicle to vehicle (Table 6, a.iv). However, a major criticism of this approach was the perceived irony of having an intervention designed to promote safety that simultaneously and actively encouraged the use of a mobile phone while driving. Indeed, a number of participants suggested that this represented a contradictory safety message and that mobile phones should therefore have nothing to do with the system (Table 6, a.v). In addition, participants raised concerns regarding the technological capacity for modern smartphones to measure the specific behaviours of interest to an eco-safe in-vehicle HMI, suggesting that additional measurement equipment would also be required. While additional sensors, or connection to the on-board diagnostics (OBD) of a vehicle, would be required to measure some eco-safe driving behaviours, statements from some participants suggested they were not fully aware of the potential capacities of their smart phones (Table 6, a.vi). Regardless of the preferred approach, a common preference among participants was for the system to operate with a high level of automaticity. That is, participants almost unanimously expressed a desire for the system to start automatically upon the commencement of a trip and be straightforward to set up. Indeed, it was argued that a less automatic system is more likely to conflict with daily life responsibilities and be associated with lower rates of prolonged usage (Table 6, a.vii). Moreover, a number of participants argued that such a system would be more appealing and useful if combined with additional

258

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

features, such as GPS, dash-cam and other in-vehicle systems (Table 6, a.viii). That is, increasing the perceived utility of the system was reported as being likely to enhance driver acceptance of the system and subsequent intentions to use. Interestingly, very few participants reported a preference for the use of head-up displays (HUD), whereby visual information and feedback is projected onto the windshield of the vehicle. A number of participants suggested that drivers are already accustomed to looking at screens within the vehicle, whereas HUD may be more distracting (Table 6, a.ix).While it remains unclear, this finding may represent a reluctance among participants to try a new technology or a lack of familiarity regarding the specific features and capabilities of this technology, however further research is required to gain a more comprehensive understanding of the issue. Overall, a high level of system automaticity appears fundamental to system acceptance and use. In addition, such an approach is more likely to increase the likelihood of long-term use and eventual adoption of use of the system as normative behaviour. Enhancing the functionality and aesthetic design of the system would also positively influence driver acceptance and subsequent rates of use. 3.5.2. Feedback characteristics Twenty participants made a total of 34 statements regarding their perceptions toward feedback characteristics of the concept eco-safe in-vehicle HMIs. The majority of participants acknowledged that there are likely to be individual differences in preferences regarding how feedback is provided by an eco-safe in-vehicle HMI and that feedback would be most effective when an individual can reliably relate to the information (Table 6, b.i). Such a finding highlights the importance for a system to be readily customizable in regards to the manner in which feedback is received. The primary debate regarding feedback characteristics centred on the most preferred presentation format of feedback – namely auditory, visual or a mixture of both. Participants varied considerably in their preferences, further highlighting the importance of providing drivers with the ability to customise their feedback preferences. That is, while some participants reported a preference for auditory feedback, others suggested a preference for visual feedback (including the use of ambient light). A number of participants even suggested that the most effective approach may be a mixture of both auditory and visual feedback, such that they complement one another to reduce driver distraction, as well as reducing the chance a driver may miss the feedback information (Table 6, b.ii). Overall, the preference for a particular approach appeared to be contingent on which feedback was perceived to be most straightforward and easy to understand, and least distractive (Table 6, b.iii). A number of participants were positive toward the use of ambient light feedback, such as a device that glows particular colours in relation to various feedback messages. Indeed, it was noted that such an approach may reduce driver distraction associated with the system by allowing them to attend to the feedback information in a largely peripheral manner (Table 6, b.iv). However, a number of participants expressed concerns regarding the distractive qualities of ambient light at night, and the emotional perception toward particular colours and the subsequent effect on acceptability (Table 6, b.v). These findings yet again highlight the importance of providing drivers with the ability to customise their feedback preferences. Taken together, participants desire an ability to customise the feedback characteristics of the eco-safe in-vehicle system to suit their specific preferences. That is, they believe the system should be flexible enough to adapt to the specific needs of each individual user in various driving contexts. 3.6. Potential problems associated with using an eco-safe in-vehicle HMI As has been mentioned in the previous sections, participants reported a number of potential problems associated with using an eco-safe in-vehicle HMI, including (a) driver distraction, (b) privacy concerns, (c) perceived accuracy and trust in the system, (d) novelty, in particular the impact of novelty on prolonged use, and (e) the promotion of unsafe and illegal driving behaviours, such as mobile phone use while driving. 3.6.1. Distraction Eleven participants made a total of 15 statements regarding the potential for use of the concept eco-safe in-vehicle HMIs to result in driver distraction. Indeed, this was the most common concern associated with the use of the system. As mentioned previously, various device types and feedback approaches were argued to have the potential to distract a driver’s attention from the driving task and thus reduce safety. Participants highlighted the irony of this risk given that the system is designed to, among other things, improve safety (Table 7, a.i). It appeared that feedback preferences among participants were heavily influenced by concerns regarding the potential distraction associated with that feedback. Specifically, a number of participants suggested that both auditory and visual feedback had the potential to be distracting or irritating to a driver and that distracting feedback may reduce the likelihood of continued use of the system (Table 7, a.ii). These findings are consistent with research which has questioned the impacts of eco-driving and other types of in-vehicle systems on driver safety (Vaezipour et al., 2015). Any additional interaction while driving may cause distraction due to increasing the complexity of the main task and therefore it may be demanding regardless of whether it is designed for eco-driving or safe driving style (Oviedo-Trespalacios et al., 2016). Therefore, these concerns are considered during the design and development of in-vehicle systems, such that an eco-safe in-vehicle system has minimal impact on driver safety. One potential approach to reducing the distracting qualities of such a system, which has been highlighted earlier, is ensuring a high level of flexibility in the available feedback preferences, such that users can readily customise such feedback according to their individual preferences and/or the driving context.

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

259

3.6.2. Privacy concerns Twelve participants made a total of 15 statements regarding privacy concerns associated with the concept eco-safe invehicle HMIs. The primary concern regarding privacy was whether data collected by the system was accessible by authorities (e.g., law enforcement, insurance companies, or government agencies) and could subsequently be used to incriminate them. Indeed, participants almost universally reported that such a feature would result in them not wanting to use the system (Table 7, b.i). According to Ghazizadeh and Lee (2014), perceptions of privacy can be negatively influenced depending on the type of data being collected and who is able to view the data. Privacy may also be a particular concern for specific groups of drivers who may be more reluctant to be monitored due to their attitudes toward technology (e.g., older drivers) or the impact that data sharing and accessibility may have on their occupational security (e.g., commercial drivers) (Hickman and Hanowski, 2011). 3.6.3. Perceived accuracy and trust in the system Seven participants made as many statements regarding their perceived trust in the concept eco-safe in-vehicle HMIs, which appeared to be heavily dependent on perceptions of the accuracy of the system assessments of their eco-safe driving behaviour. Specifically, participants appeared to be most concerned that the system would be too conservative in its measurements, resulting in drivers receiving feedback that their driving is unsafe or not eco-friendly, and that this feedback would be contradictory to their own beliefs about the status of their behaviour (Table 7, c.i). This finding has important implications for drivers who exhibit optimism biases in relation to their eco-safe driving ability. That is, such a driver may be more likely to attribute an observed disconnect between their perception of their eco-safe driving and the behavioural measurements of an HMI to conservative feedback settings, when in reality it is an accurate description of their driving performance. This highlights the need for clear communication to the driver regarding how behaviour is measured such that it is perceived to be transparent and legitimate. Overall, trust seems to be an important concept in regards to self-reported intention to use an eco-safe in-vehicle system. 3.6.4. Novelty Seven participants made a total of nine statements regarding the potential impact of novelty on continued, long-term use of the concept of eco-safe in-vehicle HMIs. Indeed, a number of participants suggested that the novelty of the system may be short-lived and that particular design characteristics, overall perceived usefulness and ease of use (e.g., automaticity) would be crucial factors influencing the likelihood of longer-term use (Table 7, d.i). In addition, three participants suggested that there may be a reduction in perceived utility of the system associated with positive behaviour, questioning what motives a driver would have to continue using the system if their behaviour no longer resulted in feedback promoting behaviour change (Table 7, d.ii). 3.6.5. Promotion of unsafe and illegal driving behaviours Two participants made a total of seven statements regarding the promotion of unsafe and illegal driving behaviours associated with the concept of eco-safe in-vehicle HMIs. As stated earlier, if a Smartphone application approach was adopted, that a mixed message would be sent regarding the dangers of mobile phone use while driving (see Section 3.5.1). In addition, the HMI may inadvertently promote unsafe driving behaviours among high-risk users, who may attempt to ‘‘get the worse score” possible. As a result, a number of participants reported that feedback must be positively-geared in order to minimize the likelihood of misuse of the system (Table 7, e.i). It may be necessary to only provide positive feedback in order to prevent driver misuse. This may be particularly relevant for younger drivers who may be more likely to abuse negative feedback elements of the system. 4. Conclusion and implications The purpose of the current study was to gather information to inform the design and development of an eco-safe invehicle HMI. This study was exploratory in nature and adopted a user-centred research approach to gain a comprehensive understanding of the specific needs of drivers, including their motivations and requirements, as well as factors that would represent limitations to use an eco-safe in-vehicle HMI. This research adds to the existing body of research given that it is, to the authors’ best knowledge, the first user-perception study of an in-vehicle HMI system designed to address both ecodriving while improving safe driving in Australia. The methodology is supported by the robust principles of the Technology Acceptance Model (TAM). A synthesis of findings is presented in Table 8. 4.1. Implications for practice Overall, this study has highlighted the potential for an eco-safe in-vehicle HMI, designed to provide information and feedback to drivers, to experience relatively high levels of driver acceptance if carefully designed with user needs in mind. We found that the perceived usefulness of an eco-safe in-vehicle HMI appears to be primarily influenced by monetary savings associated with fuel and vehicle maintenance costs. However encouragingly, improving awareness of eco-safe driving style,

260

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

Table 8 Synthesis of findings. Themes

Findings

Perceived importance of eco-safe driving behaviours

 General support of the notion that monitoring eco-safe driving behaviour is important and that an in-vehicle HMI has the potential to improve driving by increasing awareness of behaviour.  Monitoring perceived to be particularly beneficial in relation to behaviours that are difficult to subjectively monitor or that have considerable impacts on safety.  Some evidence of an illusory superiority bias, whereby individuals overestimated their skills and abilities relative to other drivers, and believe it is more important to monitor safe driving behaviours of other drivers compared to themselves.

Perceived usefulness of system a. Save fuel and money

 Strong perceptions that an in-vehicle HMI could increase the ability to monitor fuel efficiency and improve eco-driving style, most often driven by the desire to save money through reduced fuel and vehicle maintenance costs.  Some evidence that drivers believe that such a system may enhance insight and increased awareness among drivers who do not drive in an eco-friendly manner.  Strong perceptions that an in-vehicle HMI could enhance safety by increasing awareness of driving behaviour through an increased ability to monitor behaviour.  Greater levels of reported motivation to use an in-vehicle HMI among drivers who often drive with passengers, and in particular family.  Concerns expressed regarding how feedback will be received by drivers when it directly conflicts with their own perceptions of their driving ability.  Limited perceptions that an in-vehicle HMI would be useful for helping the environment. Reasons for this require more research, but may be associated with low perceptions regarding their ability to make a significant difference or paradoxical attitudes.

b. Improve safe driving

c. Help environment

Intentions to use system a. Driver type

b. Perceived ease of use/usability

c. Accessibility and cost

d. Normative influences

Design characteristics a. Device characteristics

b. Feedback characteristics

Potential problems a. Distraction b. Privacy concern c. Perceived accuracy and trust

d. Novelty

 Strong belief that particular groups of drivers would benefit more from the use of an in-vehicle HMI, such as novice drivers, parents of novice drivers, corporate vehicle fleets and as a sanctioning option for offending drivers.  Concerns expressed that more experienced drivers may be reluctant to accept feedback from an in-vehicle HMI, particularly if it contradicts their own perceptions of their driving performance.  Perceptions of an eco-safe in-vehicle HMI as being easy to use, effective and engaging observed as having a positive relationship with self-reported intentions to use such a system.  Concerns expressed regarding the potential for such a system to distract drivers from the primary task of driving and lead to cognitive overload.  Participants reported a greater intention to use an eco-safe in-vehicle HMI if it was readily accessible to them, both in terms of access and affordability.  Mixed reports regarding willingness to pay for such a system, with greater willingness reported if the financial benefits associated with using the system were perceived as likely to outweigh the costs of purchasing the system.  Financial incentives, such as rewards and prizes associated with eco-safe driving behaviour, were also reported as being positively associated with intentions to use the system.  A number of normative influences were reported that would increase likelihood to use of an eco-safe invehicle HMI, including peers, insurance companies and transport authorities.  The key element of normative influences appeared to be a trustworthy source from which evaluative information regarding the utility of the system could be obtained.  These findings highlight the potential partnership with insurance companies and transport authorities to offer incentives associated with system use and appropriate behaviour.  A well-designed, user-centric system is more likely to be associated with relatively high levels of acceptance among drivers.  Mixed views regarding the most effective approach for presenting feedback to drivers, with three preferred approaches: (i) an in-built system; (ii) a mobile system; or (iii) a smartphone application.  Strong beliefs that the device, regardless of the preferred approach, should operate with a high level of automaticity and would be more appealing and useful if combined with additional features (e.g., GPS, dash-cam).  Strong perceptions that an eco-safe in-vehicle HMI must be readily customizable in regards to the manner in which feedback is received.  Mixed views regarding the most appropriate presentation format of feedback, with support for auditory, visual, ambient light or a mixture of approaches. Preferences appeared to be largely dependent on which approach individuals perceived as being less distracting.  Distraction was the most common concern associated with the use of an eco-safe in-vehicle HMI. This finding has important implications for the design and development of any system.  Concerns expressed regarding accessibility of data collected by the system to authorities for purposes of incrimination, and the security of any data collected by the system.  Concerns expressed regarding the accuracy of the system, such as conservative measurements. This finding highlights the need for clear communication regarding how behaviour is measured such that it is perceived to be transparent and legitimate.  Some concerns expressed that the novelty of the system may be short-lived. Reported that design characteristics, perceived usefulness and ease of use would be crucial factors influencing the likelihood of longerterm use.

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

261

Table 8 (continued) Themes e. Promote unsafe/illegal behaviour

Findings  Concerns expressed that any system that incorporates a smartphone would send a mixed message regarding the dangers of mobile phone use while driving.  Concerns expressed that gamification principles and scoring may inadvertently promote unsafe driving among high-risk users who may attempt to ‘‘get the worse score” possible, highlighting the need for feedback to be positively-geared in order to minimize misuse.

and supporting eco-driving and benefits to the environment also appear to influence perceptions of usefulness, provided the system also saves the user money. We also found important individual differences regarding motivations and attitudes toward eco-safe driving. These factors can influence driver acceptance, as well as perceived usefulness of an eco-safe invehicle HMI, and therefore must be considered. The importance of considering user requirements and needs was also highlighted when investigating perceptions toward a range of design and feedback characteristics. Taken together, these findings highlight a fundamental need to develop an eco-safe in-vehicle HMI that affords users the ability to customise feedback to suit their specific preferences. That is, the system must be flexible enough to adapt to specific needs of each individual user and in various driving contexts. Moreover, a number of barriers to acceptability of eco-safe in-vehicle HMI including trust and privacy were reported by participants. Specifically, some groups of drivers may be more reluctant to be monitored due to their demographic and personality characteristics and it may be useful for the system designers to develop systems to remove or minimize these barriers. In general, distraction and the inadvertent promotion of unsafe and illegal driving behaviours are perceived to be the most likely potential problems associated with the use of an eco-safe in-vehicle HMI. Thus, impacts on driver safety should be considered during the early stages of design and development, such that an eco-safe in-vehicle HMI has minimal impact on driver safety. This study also highlighted a number of potential barriers to motivating more problematic drivers to engage with such a system. However, much like many other road safety initiatives, there are considerable gains to road safety, as well as the environment, which are readily achievable through the use of an eco-safe in-vehicle HMI by members of the driving population and therefore an inability to engage more problematic drivers would not necessarily represent ineffectiveness of such a system. So, this study has highlighted the potential usefulness of an eco-safe in-vehicle HMI for providing eco-safe advice and feedback to drivers. Moreover, it provides a number of contributions to our understanding of eco-driving in-vehicle HMIs, as well as some practical implications for designers and developers for increasing the level of driver acceptance of invehicle systems to increase driver performance. 4.2. Limitation & future research The present study is consistent with previous findings and provides evidence of the applicability of the TAM in the context of an eco-safe in-vehicle HMI. The findings suggest additional determinants to the TAM, including driver characteristics, design characteristics of in-vehicle systems and normative influences on intentions to use an eco-safe in-vehicle HMI, and future research should seek to evaluate the incorporation of these determinants into the TAM, specifically in the context of an eco-safe in-vehicle HMI. A limitation of this study is that, although data saturation was reached in the content of responses, the relatively small sample size may impact upon the generalizability of the results. In addition, the sample consists of predominately younger, urban drivers. However, given these characteristics, the sample is likely to be more experienced with technology and as such represent the target audience of a system like the one explored in this research. Nonetheless, future research will address this limitation by conducting a user study with a larger, more representative sample. Another limitation is the fact that user preferences may not necessarily reflect the most effective characteristics of an eco-safe system. Larger scale, user-centred research is required to objectively validate the findings of the current study, and represents the next phase of the broader research project from which this study is derived. Further research is also required to understand driver behaviour under technological instruction and ways to enhance driver motivation to adopt an eco-safe driving style. To explore these issues, we suggest future research should assess the use of eco-safe in-vehicle HMIs in different settings. Future research should also investigate the association between novelty and long-term use of in-vehicle HMIs to gain a clearer understanding of this relationship and identify solutions to address any issues. This study is the first stage of a larger project focused on in-vehicle HMI design and funded by an Australian Research Council Discovery Grant. The findings of this study will inform the design and development of an eco-safe in-vehicle HMI that will be evaluated in a state-of-the-art driving simulator.

262

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

Acknowledgments The authors acknowledge funding support provided by the Australian Research Council Discovery (grant number DP140102895). The first author acknowledges the support of the Queensland University of Technology (QUT), Australia to carry out her PhD. References Adell, E., 2009. Driver experience and acceptance of driver assistance system- a case of speed adaption (PhD). Lund University, Sweden. Alam, M.S., McNabola, A., 2014. A critical review and assessment of Eco-Driving policy & technology: benefits & limitations. Transp. Policy 35, 42–49. http:// dx.doi.org/10.1016/j.tranpol.2014.05.016. Barkenbus, J.N., 2010. Eco-driving: an overlooked climate change initiative. Energy Policy 38 (2), 762–769. http://dx.doi.org/10.1016/j.enpol.2009.10.021. Bennett, C., 2014. Smartphone app trial rewards safe drivers Retrieved from http://www.theherald.com.au/story/2127656/smartphone-app-trial-rewardssafe-drivers/. Braun, V., Clarke, V., 2006. Using thematic analysis in psychology. Qual. Res. Psychol. 3 (2), 77–101. http://dx.doi.org/10.1191/1478088706qp063oa. Bureau of Infrastructure Transport and Regional Economics, B., 2015. Road Trauma Australia. Retrieved from Canberra ACT: bitre.gov.au. CIECA, 2007. CIECA Eco-driving project: eco-driving in category B driver training & the driving test Retrieved from www.cieca.eu/project/. Davis, F.D., 1989. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarter. 319–340. http://dx.doi.org/ 10.2307/249008. Delhomme, P., Cristea, M., Paran, F., 2013. Self-reported frequency and perceived difficulty of adopting eco-friendly driving behavior according to gender, age, and environmental concern. Transp. Res. D: Transp. Environ. 20, 55–58. http://dx.doi.org/10.1016/j.trd.2013.02.002. Delhomme, P., Meyer, T., 2002. La recherche en Psychologie sociale: projets, méthodes et techniques. ECODRIVEN. European Ecodriving & Traffic Safety Campaigns. Retrieved from www.ecodrive.org. Ecodriver, 2011. Ecodriver EU-FP & projects Retrieved from htttp://ertico.com/ecodriver. EcoDrivingUSA. Alliance of Automobile Manufacturers. Retrieved from http://www.ecodrivingUSA.com. ECOWILL, 2014. Ecodriving Retrieved from http://www.ecodrive.org. Elder, R.W., Voas, R., Beirness, D., Shults, R.A., Sleet, D.A., Nichols, J.L., Compton, R., 2011. Effectiveness of ignition interlocks for preventing alcohol-impaired driving and alcohol-related crashes: a community guide systematic review. Am. J. Prev. Med. 40 (3), 362–376. http://dx.doi.org/10.1016/j. amepre.2010.11.012. Fagnant, D.J., Kockelman, K., 2015. Preparing a nation for autonomous vehicles: opportunities, barriers and policy recommendations. Transp. Res. A: Policy Pract. 77, 167–181. http://dx.doi.org/10.1016/j.tra.2015.04.003. Festinger, L., 1962. A Theory of Cognitive Dissonance, vol. 2. Stanford University Press. Fleiter, J.J., Watson, B.C., 2006. The speed paradox: the misalignment between driver attitudes and speeding behaviour. J. Austral. College Road Safe. 17 (2), 23–30. Ghazizadeh, M., Lee, J.D., 2014. Modeling driver acceptance: from feedback to monitoring and mentoring system. In: Stevens, A., Regan, M., Horberry, T. (Eds.), Driver Acceptance of New Technology: Theory, Measurement and Optimisation. Ashgate Publishing Ltd., pp. 51–70. Given, L.M., 2015. 100 Questions (and Answers) About Qualitative Research. SAGE Publications. Gore, A., 2006. An Inconvenient Truth: The Planetary Emergency of Global Warming and What We Can Do About It. Rodale. Graves, G., Jeffreys, I., & Roth, M., 2012. RACQ EcoDrive Research Study: final report. Retrieved from www.racq.com.au. Guest, G., Bunce, A., Johnson, L., 2006. How many interviews are enough? An experiment with data saturation and variability. Field Methods 18 (1), 59–82. Harvey, J., Thorpe, N., Fairchild, R., 2013. Attitudes towards and perceptions of eco-driving and the role of feedback systems. Ergonomics 56 (3), 507–521. http://dx.doi.org/10.1080/00140139.2012.751460. Hickman, J.S., Hanowski, R.J., 2011. Use of a video monitoring approach to reduce at-risk driving behaviors in commercial vehicle operations. Transp. Res. F: Traffic Psychol. Behav. 14 (3), 189–198. http://dx.doi.org/10.1016/j.trf.2010.11.010. Horrey, W.J., Lesch, M.F., Dainoff, M.J., Robertson, M.M., Noy, Y.I., 2012. On-board safety monitoring systems for driving: review, knowledge gaps, and framework. J. Safe. Res. 43 (1), 49–58. http://dx.doi.org/10.1016/j.jsr.2011.11.004. ISO, 2010. International Organisation for Standardization. Ergonomics of Human-System Interaction – Part 210: Human-Centered Design for Interactive Systems. ISO 9241–210. Geneva: Beuth. Jamson, S., 2010. Acceptability data - what should or could it predict?. Paper presented at the International Seminar on Acceptance, Paris, France. Jonkers, E., Wilmink, I., Nellthorp, J., Guehnemann, A., & Olstam, J., 2016. D54.1: Costs and benefits of green driving support systems. Retrieved from www. ecodriver-project.eu. Lai, W.T., 2015. The effects of eco-driving motivation, knowledge and reward intervention on fuel efficiency. Transp. Res. D: Transp. Environ. 34, 155–160. http://dx.doi.org/10.1016/j.trd.2014.10.003. Larsson, H., Ericsson, E., 2009. The effects of an acceleration advisory tool in vehicles for reduced fuel consumption and emissions. Transp. Res. D: Transp. Environ. 14 (2), 141–146. http://dx.doi.org/10.1016/j.trd.2008.11.004. Maguire, M., 2001. Methods to support human-centred design. Int. J. Hum Comput Stud. 55 (4), 587–634. http://dx.doi.org/10.1006/ijhc.2001.0503. Mark, D.F., Yardley, L., 2004. Content and Thematic Analysis. SAGE, California. Martin, E., Chan, N., & Shaheen, S. (2012). Understanding how ecodriving public education can result in reduced fuel use and greenhouse gas emissions. Paper presented at the 91st Annual Meeting of the Transportation Research Board, Washington, DC, January. Matthews, H.D., Gillett, N.P., Stott, P.A., Zickfeld, K., 2009. The proportionality of global warming to cumulative carbon emissions. Nature 459 (7248), 829– 832. http://dx.doi.org/10.1038/nature08047. McKenna, F.P., Myers, L.B., 1997. Illusory self-assessments—can they be reduced? Br. J. Psychol. 88 (1), 39–51. http://dx.doi.org/10.1111/j.2044-8295.1997. tb02619.x. Miles, M.B., Huberman, A.M., Saldana, J., 2013. Qualitative Data Analysis: A Methods Sourcebook. SAGE Publications, Incorporated. Mitsopoulos-Rubens, E., Regan, M., 2014. Measuring acceptability through questionnaires and focus groups. In: Stevens, A., Regan, M., Horberry, T. (Eds.), Driver Acceptance of New Technology: Theory, Measurement and Optimisation. Ashgate Publishing Ltd, pp. 89–104. Norman, D.A., Draper, S.W., 1986. User Centered System Design: New Perspectives on Human-Computer Interaction. Lawrence Erlbaum Associates, London. O’Reilly, M., Parker, N., 2012. ‘Unsatisfactory Saturation’: a critical exploration of the notion of saturated sample sizes in qualitative research. Qual. Res. 190– 197. http://dx.doi.org/10.1177/1468794112446106. Orfila, O., Gruyer, D., Geoffroy, D., Glaser, S., Rakotonirainy, A., Vaezipour, A., Demmel, S. (2016). Immersive driving simulation architecture to support gamified eco-driving instructions. Oviedo-Trespalacios, O., Haque, M.M., King, M., Washington, S., 2016. Understanding the impacts of mobile phone distraction on driving performance: a systematic review. Transp. Res. C: Emerg. Technol. 72, 360–380. http://dx.doi.org/10.1016/j.trc.2016.10.006. Pampel, S.M., Jamson, S.L., Hibberd, D.L., Barnard, Y., 2015. How I reduce fuel consumption: an experimental study on mental models of eco-driving. Transp. Res. C: Emerg. Technol. 58, 669–680. http://dx.doi.org/10.1016/j.trc.2015.02.005. Peden, M., Scurfield, R., Sleet, D., Mohan, D., Hyder, A.A., Jarawan, E., Mathers, C.D., 2004. World Report on Road Traffic Injury Prevention. World Health Organization, Geneva.

A. Vaezipour et al. / Transportation Research Part A 100 (2017) 247–263

263

Robb, G., Sultana, S., Ameratunga, S., Jackson, R., 2008. A systematic review of epidemiological studies investigating risk factors for work-related road traffic crashes and injuries. Injury Prevent. 14 (1), 51–58. http://dx.doi.org/10.1136/ip.2007.016766. Sauer, C., 1993. Why Information Systems Fail: A Case Study Approach. Alfred Waller Ltd., Publishers. Steinberger, F., Schroeter, R., Foth, M., Johnson, D.M., 2017. Designing gamified applications that make safe driving more engaging. In: Paper presented at the Proceedings of CHI 2017: Conference on Human Factors in Computing Systems. Stevens, A., Burnett, G., 2014. Designig In-Vehicle Technology for Usability. In: Stevens, A., Regan, M., Horberry, T. (Eds.), Driver Acceptance of New Technology: Theory, Measurement and Optimisation. Ashgate Publishing Ltd., pp. 253–267. Taylor, S., Todd, P., 1995. Assessing IT usage: the role of prior experience. MIS Quarter., 561–570 http://dx.doi.org/10.2307/249633. Vaezipour, A., Rakotonirainy, A., Haworth, N., 2015. Reviewing in-vehicle systems to improve fuel efficiency and road safety. Procedia Manufact. 3, 3192– 3199. http://dx.doi.org/10.1016/j.promfg.2015.07.869. Vaezipour, A., Rakotonirainy, A., Haworth, N., 2016. Design of a Gamified Interface to Improve Fuel Efficiency and Safe Driving. In: Marcus, A. (Ed.), DUXU 2016: Design, User Experience, and Usability: Novel User Experiences, Lecture Notes in Computer Science. Springer International Publishing, Switzerland, pp. 322–332. Van der Pas, J.W.G.M., Kessels, J., Vlassenroot, S.H., van Wee, B., 2014. The pros and cons of Intelligent Speed Adaptation as a restrictive measure for serious speed offenders. Transp. Res. A: Policy Pract. 67, 158–174. http://dx.doi.org/10.1016/j.tra.2014.07.001. Van der Voort, M., Dougherty, M.S., van Maarseveen, M., 2001. A prototype fuel-efficiency support tool. Transp. Res. C: Emerg. Technol. 9 (4), 279–296. http://dx.doi.org/10.1016/S0968-090X(00)00038-3. Venkatesh, V., Morris, M.G., Davis, G.B., Davis, F.D., 2003. User acceptance of information technology: toward a unified view. MIS Quarter. 27, 425–478. Vilimek, R., Keinath, A., 2014. User-centred design and evaluation as a prerequisite for the success of disruptive innovations: an electric vehicle case study. In: Regan, M., Horberry, T., Stevens, A. (Eds.), Driver Acceptance of New Technology: Theory, Measurement and Optimisation. Ashgate Publishing Limited, pp. 169–186. Wadud, Z., MacKenzie, D., Leiby, P., 2016. Help or hindrance? The travel, energy and carbon impacts of highly automated vehicles. Transp. Res. A: Policy Pract. 86, 1–18. http://dx.doi.org/10.1016/j.tra.2015.12.001. Walker, J.L., 2012. Research column. The Use of Saturation in Qualitative Research. Canad. J. Cardiovasc. Nurs. 22 (2). http://dx.doi.org/10.1007/s11274-0151903-5. Waylen, A.E., Horswill, M.S., Alexander, J.L., McKenna, F.P., 2004. Do expert drivers have a reduced illusion of superiority? Transport. Res. F: Traffic Psychol. Behav. 7 (4), 323–331. http://dx.doi.org/10.1016/j.trf.2004.09.009. Weber, R.P., 1990. Basic Content Analysis. SAGE. World Health Organization, W, 2014. Ambient (outdoor) air quality and health Retrieved from http://www.who.int/mediacentre/factsheets/fs313/en/. Young, M., Birrell, S., 2012. Ecological IVIS design: using EID to develop a novel in-vehicle information system. Theor. Issues Ergon. Sci. 13 (2), 225–239. http://dx.doi.org/10.1080/1463922X.2010.505270. Young, M., Birrell, S., Stanton, N., 2011. Safe driving in a green world: a review of driver performance benchmarks and technologies to support ‘smart’ driving. Appl. Ergon. 42 (4), 533–539. http://dx.doi.org/10.1016/j.apergo.2010.08.012.