IN
DEGREE PROJECT COMPUTER SCIENCE AND ENGINEERING, SECOND CYCLE, 30 CREDITS
,
STOCKHOLM SWEDEN 2021
A Concept Design for Personalized
In-vehicle Infotainment in Car
Sharing
SHUO YANG
KTH ROYAL INSTITUTE OF TECHNOLOGY
A Concept Design for
Personalized In-vehicle
Infotainment in Car Sharing
SHUO YANG
Master’s Programme, ICT Innovation, 120 credits Date: February 16, 2021
Supervisor: Pavel Karpashevich Examiner: Kristina Höök
School of Electrical Engineering and Computer Science Host company: Volvo Cars
A Concept Design for Personalized In-vehicle Infotainment in Car Sharing / En konceptdesign för personlig infotainment i bilen i bildelning
| i
Abstract
As a possible solution for environmental and social problems, car sharing is emerging. A number of studies have documented that personalization could increase convenience, comfort, and safety which becomes a shaping consumer trend, especially in cars. However, most studies in the field of personalization have only focused on privately-owned cars.
This study set out to explore the personalization in shared cars. A user-centered approach was used. Nine user interviews were conducted to explore the problems of driving an unfamiliar car. Based on the interview results, the prototype of in-vehicle infotainment was designed which can exchange data with the mobile phone and provide personalized functions. Finally, the prototype was evaluated with 5 users with the Technology Acceptance Model to test the user acceptance.
The prototype demonstrated the concept of exchanging personal information with the mobile phone. The frequently used functions on in-vehicle infotainment including calendar, navigation, and instruction were personalized in content and functionality. The evaluation results indicated that participants were possible to accept the prototype for its usefulness and ease of use but it should be noticed that privacy is an important issue when designing personalized functions. The findings presented in this thesis include the process and problems of driving an unfamiliar car, the possible approach for personal data exchange, the convenient personalized functions, and the user’s attitudes towards these personalized functions and privacy problems.
Keywords
| iii
Sammanfattning
Som en reaktion på klimatförändringar och visionen om framtidens städer ökar intresset för bildelning. Ett antal studier har påvisat hur personalising värderas högt när det gäller bilanvändning. Dessa studier har dock bara fokuserat på privatägda bilar.
Denna studie syftade till att utforska en mer personaliserad upplevelse av delade bilar med ett användarcentrerat tillvägagångssätt. Nio användarintervjuer genomfördes för att utforska problem som kan uppstå när man blir står inför att köra en okänd bil. Baserat på intervjuresultaten designades en prototyp för infotainmentsystemet tillsammans med en parkopplad mobiltelefonen och för att kunna erbjuda personaliserade funktioner till användaren. Slutligen utvärderades prototypen med 5 användare med tillvägagångsättet Technology Acceptance Model för att testa användarens acceptans.
Prototypen visade konceptet att låta bilen få tillgång till personlig information från mobiltelefonen. De mest använda funktionerna på infotainmentsystemet i bilen inklusive kalender, navigering och instruktioner personaliserades i aspekterna innehåll och funktionalitet. Utvärdering och användartest visade att deltagarna uppskattade prototypen för dess användbarhet och användarvänlighet, men det bör noteras att integritet är en viktig fråga när man utformar personaliserade funktioner. Resultaten som presenteras i denna avhandling inkluderar processen och problemen med att köra en okänd bil, ett möjligt tillvägagångssätt för utbyte av personuppgifter, design och utvärdering av personaliserade funktioner och användarens attityder till dessa samt problematisering kring integritetsfrågor.
Nyckelord
Acknowledgments | v
Acknowledgments
Firstly, I wish to thank Jingjun Ma for the great group work we have done on the project. Secondly, I want to express my gratitude to Daniel Bark, Jan Nilsson, and Alvaro Rodrigo Alonso at Volvo Cars for giving inspiring comments on the project. I would like to thank Anna Hellmark for the management of this project. In the end, I would like to thank my supervisor Pavel Karpashevich for giving detailed suggestions for the thesis.
CONTENTS | vii
Contents
1 Introduction 1 1.1 Motivation. . . 1 1.2 Research Question . . . 2 1.3 Delimitation . . . 2 2 Background 5 2.1 Car Sharing . . . 5 2.2 In-vehicle Infotainment . . . 6 2.3 Personalization . . . 72.4 Technology Acceptance Model . . . 8
3 Methods 11 3.1 User Interview . . . 11 3.2 Thematic Analysis . . . 12 3.3 Brainstorming . . . 12 3.4 Prototyping . . . 13 3.5 User Test . . . 13 4 Results 14 4.1 Interview Results . . . 14 4.1.1 Process . . . 14 4.1.2 Problems . . . 15 4.2 Brainstorming Results . . . 16 4.3 The Prototype . . . 18 4.3.1 Overall Design . . . 18 4.3.2 Functions . . . 20
4.4 User Test Results . . . 24
4.4.1 Usefulness . . . 24
4.4.2 Ease of Use . . . 26
viii | Contents
5 Discussion and Conclusions 29
LIST OF FIGURES | ix
List of Figures
4.1 Data flow between IVI and the mobile phone. . . 19
4.2 Two states of the home screen . . . 20
4.3 Two components on the home screen. . . 21
4.4 Navigation screen. . . 22
4.5 Check shared locations and contact others . . . 23
LIST OF TABLES | xi
List of Tables
3.1 Demographic information of interviewees . . . 12
3.2 Demographic information for the user test . . . 13
A.1 Interview questions . . . 43
A.2 Tasks for the User Test . . . 43
Introduction | 1
Chapter 1
Introduction
1.1
Motivation
The number of passenger cars is 330 million in Europe by 2015 [1], 270 million in China by 2020 and 273 million in the United States by 2018 [2]. The primary concerns of a large number of cars are rising private car ownership costs and greenhouse emissions [3]. It has previously been observed that the fixed cost can’t be ignored as people only use their cars for a short time each day [4]. Recent evidence suggests that car usage is the major source of air pollution, noise pollution, and carbon dioxide [5].
Car sharing is becoming popular around the world [6, 5, 7]. It is an emerging industry that is considered as a possible solution for the problems mentioned above. Besides, it has several benefits including increasing mobility, equity, and economic development [8]. It is a service that users can access a shared fleet of vehicles which enables people to use a car when they need it without owning it [5]. Research in this area has shown that car sharing could reduce the cost [9] and an average car sharing user can reduce carbon dioxide emissions by 39% to 54% [10].
Personalization in cars becomes a shaping consumer trend [11]. It has been noted that personalization could enhance usability [12] and increasing convenience, comfort, and safety [13]. In-vehicle infotainment (IVI) is an embedded system that provides many information and entertainment services, including vehicle-specific information, navigation, and TV [14]. Researchers aim to provide a personalized experience for drivers by customizing the user interface of IVI [15, 16] and using a platform-independent personal profile [17].
2 | Introduction
developed solutions for the mobile phone’s integration in cars [19,20]. These solutions allow drivers to use apps running on the phone, including music, podcast, navigation, voice assistant, and calendar. In this way, drivers can personalize the car with their own devices and the phone can provide network connectivity to the car [21].
However, previous studies of personalization have only been carried out in privately owned cars. Very few studies have explored these topics in shared cars. Car sharing users will drive a different car every time. What is not yet clear is how to achieve personalization with the integration of the mobile phone in shared cars. If this problem can be solved, people will be more willing to try the less-cost and environment-friendly car sharing service.
This research cooperates with Volvo Car Corporation ∗. Volvo Car Corporation aims to become a global and diversified mobility service provider [22,23]. This work will generate fresh insights into the IVI design of shared cars for Volvo and other vehicle manufacturers.
1.2
Research Question
The primary aim of this thesis is to explore the personalization in shared cars with the mobile phone integrated. This research seeks to address the following question: how to make the driving experience in shared cars personalized with mobile phone’s integration?
To answer the research question, this thesis followed the approach of user-centered design [24]. The study used interviews with thematic analysis [25] to figure out the process of driving an unfamiliar car. Then the concept IVI design to provide personalized functions is designed according to the interview results. Finally, a user test is conducted with the approach of Anderson [26] to examine the user’s acceptance of the proposed prototype.
1.3
Delimitation
Introduction | 3
Background | 5
Chapter 2
Background
2.1
Car Sharing
The Sefage program in 1948 was the first implementation of car sharing [30] but it had not gotten much attention until the 1990s because of the rising oil price and congested traffic network [31]. After that, fifty car sharing programs have been deployed in North America since 1994 [32]. A recent study has indicated that car sharing has expanded to approximately 1,100 cities worldwide in October 2010 [9]. The number of car sharing users is growing rapidly as the research data have indicated that it is 7 million by 2015 and is expected to be 36 million by 2025 [33].
This thesis focuses on free-floating car sharing. To illustrate the features of free-floating car sharing, SHARE NOW [34] is taken as an example. It is a popular free-floating car sharing service that is operated in 16 major cities in 8 countries in Europe. Users should use the SHARE NOW app to find an available car near you, reserve it and walk to it. Users can drive the car whenever they like as there is no car rental station. Then the car can be unlocked with a push of a button in the SHARE NOW app. Finally, users can simply drop off anywhere.
6 | Background
2.2
In-vehicle Infotainment
In-vehicle infotainment (IVI) is an embedded system that provides information and entertainment services [14]. Previous research has established several design implications for IVI including adapting the IVI system to the driving context, minimizing interaction by sharing information between services and devices, defining a consistent non-visual interaction model for IVI systems, and designing IVI systems for multiple users and social interaction [37]. These implications about interaction are valuable when designing IVI. However, the implications are proposed for privately-owned cars and it requires more research to figure out whether they are suitable for shared cars.
A number of studies have begun to examine the interaction with IVI. A study has found that the interaction style of touchscreen IVI is similar to the mobile phone [38]. Chalmers [39] has indicated that the quality of interaction could be improved by letting each tool be itself. This result suggests the interaction of IVI and the mobile phone doesn’t need to be the same as the IVI usually has a bigger screen. For basic interactions, it is suggested that page-by-page scrolling is better than kinetic scrolling for less distraction [40]. As for search, it has been observed that the search-based approach is better than the menu-based approach [41].
Up to now, several studies have indicated non-touchscreen interaction methods can reduce the amount of time a driver spends looking at IVI [42]. The voice interaction is the only wide-band communication channel that is not engaged in driving [43]. However, the problems of voice interaction can’t be ignored such as the specific words for triggering commands. A study has found more problems including car noise interference, a limited number of supporting commands, and an unclear menu structure [44]. Surveys such as that conducted by Kim [45] have shown that users with more prior experience of using voice interface from other smart devices are more likely to change to the voice user interface in IVI.
Background | 7
2.3
Personalization
In the past, personalization referred to a system that can automatically adjust its behavior and interaction to suit the user’s needs [46]. In a more recent study, Blom [47] has given another definition to personalization as
"a process that changes functionality, interface, information content, or distinctiveness of a system to increase its personal relevance to an individual."
Blom mentioned two approaches: user-initiated personalization and system initiated personalization. User-initiated personalization is to describe adjustments initiated by the user to achieve a desired goal such as tuning the climate system. System initiated personalization is a system that uses a user profile to provide personalized content. For example, e-commerce websites recommend goods based on shopping history.
A study has shown the advantages of personalization including increasing convenience, comfort, and safety [13]. It is also an effective way of suiting the difference between individuals and can enhance usability [12].
However, the controversy between personalization and privacy never stops. Some works indicate that privacy is more important than personalization benefits [48] while some works have the opposite results [49]. To further investigate the paradox of personalization and privacy, Sutanto et al. [50] find users trust in applications that don’t send personal information to third parties. This finding has suggested telling users how their privacy is processed is important.
To examine how users perceive personalization in car sharing, Bardi et al. [6] interviewed Zipcar∗ users and found shared car users don’t engage in personalizing the car, e.g. seats and radio settings. Then he claimed that the anonymity of consumption is important in car sharing. However, it should be noticed that this result was from several Zipcar users in the U.S. and maybe not applicable to other car sharing services in other countries.
8 | Background
personal settings should be gathered and overviewed easily by including these settings in a separate menu.
However, Ericsson’s study mainly focused on user-initiated personalization and didn’t investigate the IVI much. Hasenjager [51] has indicated that IVI is the main target of personalization. For instance, Letchner et al. [52] create a route planner that can use the driver’s past GPS logs to generate personalized routes.
A previous study has indicated context is used to improve personalization [53] and can decrease the need to manually performing tasks or retrieving information [37]. Many researchers have investigated how to use context to provide a better experience. A study has summarized the context can be location, identity, time, history, other people, and much other information in the environment [54]. In a design project, an audio content recommendation system has been proposed based on location and driver states [55]. Hasenjager [51] has designed a system that can give information during driver’s downtime, for instance, waiting for the red light. Garzon [56] has proposed the contextual personalized approaches which are prioritizing functions based on the use of frequency and rearranging menu structure according to context. Overall, these studies highlight the need of using context to improve personalized functions. However, there has been very little related research about personalization in shared cars. Kuemmerling [17] has researched this issue and found out 94% of people are willing to use a technology that allows the faster configuration of temporarily used cars in a survey. Then Kuemmerling proposed the concept of an independent mobility profile which can be transferred to a car and enables the automatic adjustment of settings in the car. This profile consists of four data categories: general information about the driver, information about preferred vehicle settings, a list that saves hardware-dependent settings for every car that has been used, and a digital logbook that is updated when leaving the car with information that would normally remain in the vehicle’s storage. This study suggests that taking the profile and the car apart is a possible approach to solving the problem of driving different cars every time in car sharing.
2.4
Technology Acceptance Model
Background | 9
using a particular system would be free of effort. Whether a user will use a system is determined by the attitude towards using. And the attitude is affected by PU and PEOU. Besides, PEOU has a direct influence on PU.
A systematic literature review concluded that TAM has been applied to different information systems under various situations and its robustness has been proved [59]. Thus far, a number of studies have used TAM to evaluate the personalization in different areas including E-government service [60] and mobile health service [61]. TAM is also widely used in research about cars to test the driver’s acceptance of in-car technologies (e.g. navigation system [62]) and car sharing concept [63]. The study from Osswald et al. [64] provides new factors for TAM including performance expectancy, effort expectancy, anxiety, self efficacy, facilitating conditions, perceived safety, and attitudes.
Methods | 11
Chapter 3
Methods
The user-centered design approach consists of concept generation and ideation, prototype design, and evaluation [24]. It was used in the following steps. Potential car sharing users were interviewed to understand the problems and user needs. A brainstorming was conducted with a researcher at Volvo Car Corporation to generate ideas for solving the problems found in interviews. The prototype was created according to the results of brainstorming. Finally, the user test was conducted in the framework of TAM to test the user’s acceptance of the proposed IVI.
3.1
User Interview
The aim of user interviews is to find out the current experience of driving an unfamiliar car.
The interviews were conducted in a semi-structured approach. The benefit is that the interview results will be in a similar form which is easier for further analysis. Another benefit is that additional questions can be asked to clarify the details. The approach from Spradley [66] was used by asking general probing questions and then asking for additional information. The general probing questions were listed in AppendixA.1.
The criteria for selecting interviewees was the experience of car rental or car exchange. Free-floating car sharing is not popular in Sweden because it is innovative and not operated in many countries. Nine interviewees (see Table
12 | Methods
Interviewee Gender Age Nationality Occupation Driving Experience
1 M 25-30 Sweden student 6 years
2 M 25-30 China student 5 years
3 F 25-30 China student 5 years
4 M 25-30 India engineer 6 years
5 M 25-30 India student 10 years
6 M 25-30 Sweden student 10 years
7 M 25-30 India student 3 years
8 M 25-30 Sweden student 7 years
9 M 25-30 China student 3 years
Table 3.1: Demographic information of interviewees
3.2
Thematic Analysis
To analyze the qualitative result, thematic analysis [25] was adopted. This method is particularly useful in analyzing interview results for its flexibility [67].
Thematic analysis was conducted in an inductive way for interview result since car sharing didn’t have a firm theoretical framework. Codes were summarized or directly quoted from an interviewee’s sentence.
For the user test result, thematic analysis was based on TAM [57]. Codes related to usefulness, ease of use, and attitudes were extracted.
3.3
Brainstorming
The purpose of the brainstorming session was to obtain ideas for solving the problems found in the interview. The brainstorming was conducted with a mixed approach of individual brainstorming and group brainstorming because individual brainstorming can generate more ideas than group brainstorming [68].
Methods | 13
3.4
Prototyping
Framer [69] is a web-based prototyping tool and it was used to create the IVI prototype. In order to concentrate on the topic of personalization, only related functions and visual elements were designed. For less relevant functions, like music and climate system, simple user interface elements were created as placeholders without further interaction.
3.5
User Test
Many researchers have utilized TAM to measure user acceptance in a qualitative approach [70, 26]. Anderson [26] uses semi-structured interviews with questions based on TAM and then conducts the thematic analysis to evaluate the user experience of mobile health apps.
The user test was conducted with a similar approach from Anderson [26]. To begin the user test, participants were told the background of the thesis and instructed to follow the think-aloud protocol. Then participants were asked to finish the tasks described in AppendixA.2. Finally, some questions (see Appendix A.3) were asked according to the original TAM [57]. These questions were designed to collect the user’s opinions about the proposed functions, the moments when the users were confused, and the participant’s overall attitudes about the proposed IVI.
Five participants (see Table3.2) were recruited. All of the participants had sufficient driving experience. Four of the participants were young students. The participants were similar to the typical car sharing users who are male, young, highly educated with high income levels [35, 36]. The number of participants was five as the ratio of benefits to cost is the greatest when there are three to six test users [71].
The user test was conducted indoors due to COVID-19. The prototype worked on a computer screen and participants could interact with it using a mouse.
User Gender Occupation Driving Experience
1 F student 5 years
2 M programmer 10 years
3 M student 3 years
4 M student 8 years
5 M student 8 years
14 | Results
Chapter 4
Results
4.1
Interview Results
The purpose of user interviews is to investigate the process of driving an unfamiliar car and find problems in the process by asking interviewees about their previous car rental or exchange experience. The scenario was usually trip as a majority of interviewees (78%) indicated that they used the car for a trip.
4.1.1
Process
It was found that the process was similar for different interviewees. First, all the interviewees spent one or two minutes adjusting the seats and mirrors. Second, all of the interviewees personalized the car by integrating the phone to IVI to play music or navigate. Third, all the interviewees adapted the car but the adaptation time varied, which means interviewees who had much previous driving experience could adapt more quickly.
Adjustments
After getting on the car, all of the interviewees spent one or two minutes adjusting the car’s hardware settings including seats and mirrors. The adjusting process was similar for different interviewees as one interviewee said below.
"So I just sit down, adjusted the seat. And then I looked around. And I just started (the engine), looked around, looked in the mirrors to see if I got to adjust them. And then I drove off."[User
Results | 15
Using IVI
All of the nine interviewees reported that they connected the mobile phone to IVI for music and navigation. All of the interviewees played music with the car’s sound system while the music was saved on the phone. The reason was simple as one interviewee added: "The speaker in the car is good. (...) Music
is an important part."[User 2]
Most interviewees (8 out of 9) used a navigation function while driving whether on a phone or on IVI. Three interviewees indicated that they used map apps on the phone because the map data was up-to-date. Five interviewees stated that they used the navigation function of IVI for its big screen was easy to view. For example, as one interviewee commented:
"There is no reason to use the small screen if I have a big one. (...) It is convenient to glance at the big screen during driving."[User
2] Adaptation
The adaptation time varied from a couple of minutes to an hour according to the results. Interviewees who had much previous driving experience thought the adaptation was not difficult and could adapt more quickly because of similar basic operations. As one interviewee said: "There is no difference
in basic operations. (...) The cars are all the same."[User 2]
Interviewees with less driving experience had some uncertainty when driving. As one interviewee put it:
"You’re not sure how to do all the common things. (...) So you have this a little bit of fear because you don’t want to have an accident."[User 1]
4.1.2
Problems
Data IssuesAlthough IVI had several advantages for its big screen and better sound system, the comments below suggested that the problems were the lack of data and the outdated data. The car which was not owned by the interviewee didn’t have favorite music and saved locations. The data on IVI, especially the navigation data, was out-dated.
16 | Results
"The navigation system on the car is just a reference because it is not updated in time as the phones."[User 3]
Unfamiliarity
When starting driving, most interviewees (7 out of 9) reported that they had a feeling of unfamiliarity with the car, traffic situation, and driving regulations. One interviewee added: "I think the instructions before (driving different
cars) are good. So I know I’m sort of more familiar."[User 1] The result
suggests that practice and instructions could make drivers more familiar with the car.
The interviews also showed that the difference between cars includes driving feelings and operations details. The driving feeling includes operating the accelerator, brakes, and wheels. Drivers can only get familiar with it by practice. The operation details include symbols, button locations, and the process of performing a task.
"When I switch a car from the B92 to S60, I’d say that was a big difference in terms of acceleration and how it felt on the road."[User 8]
"My perception is that the symbols are not standardized, actually. So the user might not know the symbol."[User 5]
"You need to look at the details again, like how to answer the phone, how to connect Bluetooth, and how to set the climate system."[User 2]
Privacy Issues
Concerns were expressed about privacy as interviewees were careful when sharing privacy on other cars. For example, one interviewee said:
"I had to accept the connection with my phone. And then the car asked me to share like contacts and messages and stuff like that, which I did not want to do since it was not my car."[User 8]
4.2
Brainstorming Results
Results | 17
Data Source
To solve the data issues, a possible solution is similar to current products like Carplay which uses the phone as a data source and computing device. However, this idea doesn’t make full use of IVI. IVI just works as the screen and the computing power is wasted.
Another idea is to use an independent data source and use IVI for computing. The IVI should be able to exchange data with the data source and then functions of IVI can be personalized in information content and functionality by using these data.
The possible options for data sources include the online account and the mobile phone. If the shared cars have internet connectivity, the cars can access user’s preferences and personal data saved on the user’s online account. Although an online account seems more convenient for users, it requires Internet connectivity. Not every place is currently covered by internet connectivity. So this approach may not work in some areas which are not covered by internet connectivity.
Most free-floating car sharing services are based on mobile apps [29], it’s natural to think up the idea of using mobile phones. People are accustomed to using the mobile phone and saving various personal data in different apps, including favorite music, saved locations, calendar, and contacts. For instance, the calendar on IVI can use the data of the calendar application on the phone, display personalized calendar events, and provide other features like calling others if the necessary information is provided.
Solutions for Unfamiliarity
Autonomous driving can be a solution to the problem of unfamiliarity as the car sharing users don’t need to drive the car themselves. However, this technology is still in development and far away from commercial use. So this idea is not selected.
From the interview results, it is found that instructions could assist the adaptation process and make the drivers more familiar with the car. Several ideas about the design of instructions were discussed.
The instruction should be relevant to the car model and specific so users can better understand them. The type of instructions can be animation, pictures, and text.
18 | Results
on the user’s driving history. For instance, if the driver has used this car model several times, the instructions of the car usage should be given lower priority. The developed idea is that the instructions should dynamically display instructions according to context and the user’s driving history.
There is an idea to include parameters of the car performance in the instructions. The car performance includes maximum speed, acceleration time, braking distance, etc. These parameters can be quantified by comparing to the car which the driver used last time or common objects. However, it is thought that these parameters are not necessary when driving and drivers can learn these parameters by practice. So this idea is not selected.
There is another idea of giving instructions in a gamified way. Users need to go through a game level that giving instructions about car usage. However, this idea is time-consuming and the instructions are fixed in the game so it is hard to suit the needs of different drivers.
Privacy Settings
For privacy issues, users should be given the ability to choose what kind of private information to share with IVI. Therefore, a function of privacy settings should be included. There are different ideas about how to design the privacy settings. The first idea is to manage privacy settings only on the phone. Users should set the privacy settings before using the shared car. However, users may change their minds when getting on the car. So the selected idea is to provide privacy settings in IVI and the mobile phone.
Then there are some factors to consider. For the time of asking permissions, it can be the time of reserving the car or the time of getting on the car. As the time reserving the car is too early, it is decided to use the time of getting on the car. For the display style, if using popup windows, there will be too many popups when the drivers use IVI. So the idea of using switches is selected. For the display location of requests, the idea is to show all the requests for personal information on the first screen of IVI. The problem is that it is difficult for users to understand why this information is needed. This idea is improved by showing requests within relative functions. Users can better understand the reason why this personal information is needed.
4.3
The Prototype
4.3.1
Overall Design
Results | 19
personal data with the mobile phone. The proposed IVI is an independent system that can exchange data with the phone via the car sharing app, which works as the bridge of data exchange (see Figure4.1).
If users are willing to share privacy and access the personalized functions, the personal data in various apps on the phone (including calendars, To-Do lists, maps, and other apps) can be transferred to IVI. Otherwise, the users can still use basic IVI without personalized features. The data generated from the car could also be transferred to and saved in the car sharing app for future use such as more detailed driving history.
The purposes of the prototype are to show the personalized concept and to test the user acceptance of it by letting users use it and perform tasks in a trip scenario. With the prototype, car sharing users can get a personalized experience by accessing personal data even in different cars. Five frequently used functions are designed with personalized features including privacy settings, calendar, navigation, instruction, and voice assistant. Detailed descriptions of these functions are listed in the following.
20 | Results
4.3.2
Functions
Privacy SettingsWhen users get on the car, the IVI is in standby mode. When users are ready to use the IVI and connect their phones, users will see the privacy settings if it is the first time when they use the car sharing service. The function of privacy settings (Figure 4.2(a)) is designed to show the requests for permissions in related functional areas to assist users to understand why these permissions are needed. If users want to change the permissions, they can go to the settings of IVI and modify their privacy preference.
After giving permissions, the car sharing app will begin to collect necessary personal data, transfer the data to IVI, and then the functions can be personalized respectively. If not given permissions, these functions will still work without personalization. Personal data that have been transferred to IVI will be used locally and won’t send to third parties to protect user’s privacy from other companies. These data will be deleted when finishing driving to prevent the following driver of the same car from accessing these data.
(a) Home screen with privacy settings. (b) Normal home screen. Figure 4.2: Two states of the home screen
Calendar
Results | 21
The information content of the calendar function is personalized. Users can easily check their schedules and To-Do lists on IVI and edit them with the voice assistant. The changed part will be transferred back to the user’s phone with the assistance of the car sharing app.
The calendar function provides personalized features including contacting others and starting navigation when the contact and location information is available. If the calendar events contain locations, there will be a navigation button. For instance, both events in Figure 4.3(a)) have a navigation button. The locations can be geographic locations (e.g. the mountain) or locations shared with the user (e.g. Alice’s location). When the button is clicked, IVI will go to the navigation screen and set the location as the destination. If the events contain the name of the contact, a call button will show up. The car sharing app will search the name in the user’s contact to find the phone number and then transfer it to the calendar on IVI. By clicking it, the user can start a call with earphones or the speaker in the car.
(a) Calendar component is on
the top left of home screen. (b) Navigation component ison the top right of home screen.
Figure 4.3: Two components on the home screen. Navigation
22 | Results
The navigation screen (Figure4.4) can also be accessed by clicking the map area or the search button. In the panel, the three buttons in the first line are as same as the buttons in the navigation component on the home screen. In the second line, there are three different categories. Users can check locations in different categories and click to set as the destination. The calendar category displays all the locations mentioned in the calendar events by checking the user’s calendar data. The favorite category shows locations that the user has marked in map apps on the phone with these locations transferred to IVI by the car sharing app. The history category shows locations that have been set as destinations before by referring to the driving history in the car sharing app.
Figure 4.4: Navigation screen.
Results | 23
(a) Locations are displayed as
pins. (b) Multiple contact methods
Figure 4.5: Check shared locations and contact others Instruction
The instruction function (bottom left of Figure4.2(b)) is a solution for adapting to drive a different car caused by the difference of operation details. This function is based on the the contextual personalized approach using frequency and the context [56]. This function introduces the hardware, features, and driving regulations according to context and personal data.
First, the instructions are selected according to the context including time, weather, and location. The context data is from car sharing app which can read time and weather from other apps and get a current location with GPS. The concept of how to use the context is described in Figure4.6(a). Then, this function uses system initiated personalization by giving each instruction item a different priority level based on the user’s driving history and the history of referring instructions. If the user has driven the same car model several times, the instructions related to the operation will be given low priority. If the user usually refers to the instruction for the same component, the instruction about the component will be given high priority. Besides, the instruction related to safe driving will be given high priority. In the end, high priority instructions are displayed on the home screen.
24 | Results
(a) Framework (b) Instruction Details Figure 4.6: Instruction function.
Voice Assistant
Considering the advantages of voice interaction stated in Section2.2, users can find it in the middle right in the home screen (Figure4.2(b)) and the bottom left in the navigation screen (Figure4.5).
The voice assistant is the hub of all the personalized functions mentioned above by supporting the users to use these functions with voice. The usage is simple that users can speak out the commands and then the voice assistant would do what they want.
4.4
User Test Results
The results were analyzed with the factors of TAM including usefulness, ease of use, and attitudes.
4.4.1
Usefulness
The usefulness theme covers the user’s opinions about personalized functions. Personalized functions are convenient
Four out of five users gave positive feedback to the functions clearly. One user acknowledged that the calendar and navigation functions were convenient and could help her save time because the information is provided automatically when necessary. As the user said: "These functions are great. I don’t need to
Results | 25
The calendar can be improved as two users mentioned that it would be better to divide the calendar into the office calendar and private calendar for the car was a public place.
Another user indicated that all of the functions were useful but it depended on the scenario. As the user explained: "For example, music is not important
so much but it depends on the distance of commuting."[User 4]
It was noticed that one user perceived the buttons in the navigation component differently. The user didn’t realize the buttons were personalized and thought these buttons were used for searching nearby locations. As the user said:
"It’s the first time I used it. I thought the system is independent. I didn’t set these locations. If I know it can read my Google maps data, I will use it."[User 4]
Unnoticed instruction function
When asked to find the wiper instruction, all of the users could locate the correct place on the home screen. However, most of the users just finished the task and did not mention this function in the following interview. Only one user commented this function was good when the concept of the instruction function was introduced.
Voice assistant is convenient
Three out of five users felt the voice assistance was convenient. One user liked the voice assistant very much and tried to finish many tasks with it including setting temperature and finding the wiper instruction. As the user commented:
"I like the voice assistant. It’s convenient if it can work well in a real situation. For example, it will be great if I can tell the voice assistant to set the temperature to 26."[User 3]
Unwillingness of sharing contact
A problem was found that two users still chose to call someone via the phone instead of using the call button in the calendar component. The two users then explained that it was because they were unwilling to allow the IVI to read the contact information. As one user said:
26 | Results
At the same time, the two users gave permission for IVI to access the location because they thought it was reasonable to use the location for navigation. The result suggested users had different feelings about different kinds of privacy information and users were careful about the contact data.
4.4.2
Ease of Use
In the theme of ease of use, results about usability were reported.
One user thought the user interface of IVI was “pretty straightforward” but the user added that it would be better if the home screen was editable. As the user indicated:
"I don’t need some information all the time and I think most drivers want to choose what they see."[User 2]
The prototype still had some usability problems. Considering driving at night, the background color was set to dark colors because light colors would be too distracted but one user was not satisfied with the background color of the map as he said: "I think the background color of the map should be light.
And I usually listen to the instructions."[User 3]
For the location categories in the navigation screen, one user indicated that she would search instead of choosing from the categories because of a large number of marked locations.
Two users felt the music module was too large on the navigation screen. As one user said:
"I want everything else to disappear when I use navigation. The music part is too big. It can be smaller."[User 3]
Most cars now have a two-zone climate system but 4 out of 5 users were confused about the climate control area because they didn’t know why there were two numbers about the temperature.
"I don’t understand why there are two 26 numbers. I usually think one number is temperature and one is humidity".[User 4]
4.4.3
Attitudes
The attitudes towards the acceptance of IVI had a tight relationship with privacy. Like many other personalized functions, privacy was mentioned by most of the users.
Results | 27
including recording and camera. One user trusted big companies but was concerned about privacy that may be left in the shared cars.
"I am not worried because Volvo is a big company. (...) I am concerned about the permission of the calendar because it’s a shared car but it’s reasonable to use the location."[User 3]
However, other users held different opinions. One user was suspected about the functions which ask for many permissions. As the user said:
"It asks for much authorization. The functions I get are not matched with the permissions I give."[User 5]
The user also indicated that before giving permissions, why these permissions were needed was not clear and more information was needed to assess the necessity of sharing privacy to IVI. As the user commented:
"The convenience from the personalization is not compared with the complexity of the privacy settings, in other words, the worry about privacy."[User 5]
Another user was worried about the leakage of location information. As the user said:
"I am worried that other people know about my location. (...) Companies will get everyone’s data silently. The only way to protect it is the law."[User 1]
When this user was asked whether the design of privacy settings could make her less worry about privacy, the user said:
Discussion and Conclusions | 29
Chapter 5
Discussion and Conclusions
5.1
Discussion
Achieve Personalization with the Mobile Phone
Personal data are essential in creating a personalized experience. However, personal data can not be saved in shared cars as the data will be lost when the user returns the car. The solution to this problem is using the phone as the data source of IVI which is similar to Kuemmerling’s idea of separating the user profile and the temporarily used cars [17]. While Kuemmerling’s idea has focused on the hardware settings [17], this design concentrates on personal information. This thesis proposes a relatively new approach. With the car sharing app working as the bridge, IVI can exchange personal data with many other apps on the user’s mobile phone to get information about calendars, contacts, driving history, and locations and transfer driving and instruction referring history back to the phone to assure that data are always saved in user’s mobile phone. In current car sharing, the phone is used for finding and unlocking cars. This approach adds the function of the data source to the phone which is the basis of personalized experience. Several personalized functions including the calendar, navigation, and instruction are proposed in the prototype. With the mobile phone’s integration, these functions can change the functionality, interface, and information content based on personal data. The user test results suggested that most of the users found this design useful. It can be therefore assumed that this design resolves the key problem of personalized car sharing that users will drive different cars. However, it should be noticed that the user test had limited participants and the environment was not similar to driving.
30 | Discussion and Conclusions
the design of recommending audio content based on location and driver states [55] and the design of prioritizing functions based on the use of frequency [56]. The participants in the user test thought these functions are convenient which indicated that personalized functions can simplify the process of performing tasks in the driving scenario. Hence, these personalized functions can be the start points for future personalized IVI design by introducing the idea of personalizing functionality, interface, and information content of IVI functions.
User Acceptance of Personalization
One of the aims of this study is to assess the acceptance of the proposed IVI. The results of the user test show that participants had positive opinions on the factors of usefulness and ease of use as they thought the prototype was convenient and straightforward. According to TAM [57], the possibility of accepting the proposed IVI was high. This result is consistent with that most of the users are willing to use personalized technology for fast configuration in temporarily used cars [17].
The most important result is that most of the participants thought the prototype was convenient. This result confirms that personalization can increase convenience [13]. Several factors could explain this result. First, information is automatically fulfilled. For instance, users don’t need to input locations in the navigation function. Second, functions are varied to meet the user’s needs according to personal data, e.g., the contact function in calendar and navigation. This result proves that the prototype successfully explores the personalized design of IVI. However, this result should be interpreted with caution considering the limited number of participants and the indoor test environment which is not matched with the real driving environment.
Issues in Personalization
Discussion and Conclusions | 31
functions are not suitable to test in a relatively stable prototype. A further study with the developed version of the instruction function is therefore suggested.
As Heikkinen suggested in the IVI design implication, managing privacy should be provided [37]. In this study, privacy settings were designed. It met the need that users wanted to manage privacy individually. However, there is a problem that the guidance for changing the settings is missed. The user test results suggested some users would change their minds about sharing personal information. The introduction should be provided. It was a surprising result that the users were still careful when sharing privacy. This result supports the claim from Bardhi et al. [6] that users pay attention to those actions may leave hints in shared cars. Freidman et al. [72] have suggested that the majority of people maintain some privacy in public and the consideration involves informed consent. This result is likely to be related to the public property of shared cars and the missed information of the requests. For example, the requests don’t include all the permissions the function needs. So the privacy settings can be improved by giving more information to the user.
Privacy issues affect user’s acceptance as Sutanto [50] addressed. Participants didn’t have a common view of privacy. Some participants were not worried while some participants were suspected of the permissions asked for personalized functions. Besides, some participants were worried about car sharing company’s massive data collection. This accords with previous research that has different results of the paradox of personalization and privacy [48, 49]. The relative parties of privacy leakage can be the following user of the same, the car sharing company, and the mobile phone’s producer. The data on IVI will be deleted after driving so the following user of the same car can not get accessed to it. The car sharing company should be restricted by law and the data collection should be under the regulations. As for the mobile phone’s producer, these companies usually have strict data protection rules. Users have decided whether to share personal information with these companies before using the car sharing service. If users don’t save personal information on the phone, users can use the shared cars without personalized functions. It’s up to the users.
One participant mentioned it was possible to change the idea of sharing personal information. Although the privacy settings can be accessed again in the IVI settings, the prototype doesn’t give clear instructions about how to access the privacy settings. This should be improved.
Importance of Voice Assistance
32 | Discussion and Conclusions
functions. It is an interesting result that the majority of participants in the user test were willing to use it. The user test results also indicated that voice interaction can substitute touch screen interaction in many aspects including editing information, setting a climate system, and contacting other people. As car sharing service requires the user to have a mobile phone, the participants had a high possibility that they have used the voice assistant before. This finding is consistent with that of Kim [45] who found users who have used the voice assistant on other devices are more likely to use it on the car. It can therefore be assumed that voice assistant can play a more important role in the interaction with IVI for its high acceptance, ease of use, and safety.
5.2
Limitations
The obvious limitation is that the test environment was not consistent with the real driving situation. Due to COVID-19, the test was not conducted in the car environment but in the indoor environment. The participants didn’t need to pay attention to driving. As a result, car-related factors were difficult to evaluate in the indoor test. The car-related factors include perceived safety and anxiety [64] and these factors are important in evaluating the prototype.
The variety of the interviewees and participants is not sufficient which may affect the quality of user interviews and user tests. The interviewees and the participants in the user test were mainly male, students and in a similar age group. Although the background of the interviewees and participants are close to the typical profile of car sharing users [35, 36], there are many different backgrounds which are not included in this study, for example, the elderly, the disabled, and users living in the countryside.
The prototype is not tested in other scenarios. The test scenario is a trip scenario based on the interview. The personalized functions were tested in the trip scenario. However, there are many other scenarios of car sharing. For example, commuting and shopping scenarios are a very common scenario [73]. Whether these functions can work well requires further research.
5.3
Future work
Discussion and Conclusions | 33
In this study, the user test is not conducted in the car. So the factor of safety, which is important in the driving scenario, could not be measured. Some laboratory experiments with a driving simulator can be designed to test the prototype for more factors. Then the results will be more accurate.
5.4
Conclusions
This study set out to explore personalized experiences in shared cars. To provide a personalized experience, the concept design of the IVI prototype is proposed. The prototype can exchange personal data with the user’s mobile phone while the car sharing app works as the bridge. The functions of the prototype are personalized in information content and functionality with personal data. The events in the calendar function and locations in the navigation function are personalized by using the data saved on the user’s phone. More features including quick contact and quick navigation inside the two functions are provided to users when necessary personal data are available. For less experienced drivers, the function of personalized and contextual instructions can help them become familiar with the car. Besides, the privacy settings allow users to manage privacy and the voice assistant is the hub of personalized functions.
The user test has shown positive results in the aspects of usefulness and ease of use. These results could lead to a high acceptance of the prototype according to TAM but the instruction function was not fully tested as participants didn’t realize it was contextual. The results have also indicated the importance of voice assistants and suggested making better use of them in the future. However, there are some issues with privacy. It can be concluded that users care about their privacy to different degrees and the degrees are also related to the type of privacy. In general, the findings of this study suggest that the approach of exchanging data with mobile phones and the proposed functions can provide users with a personalized experience.
34 | Discussion and Conclusions
REFERENCES | 35
References
[1] O. I. des Constructeurs d’Automobiles, “Passengers cars world vehicles in use,” 2015.
[2] List of countries by vehicles per capita. [Online]. Available: https: //en.wikipedia.org/wiki/List_of_countries_by_vehicles_per_capita
[3] U. Nations, “68% of the world population projected
to live in urban areas by 2050, says un.” [Online].
Available: https://www.un.org/development/desa/en/news/population/ 2018-revision-of-world-urbanization-prospects.html
[4] S. Shaheen, D. Sperling, and C. Wagner, “Carsharing in europe and north american: past, present, and future,” 1998.
[5] R. Katzev, “Car sharing: A new approach to urban transportation problems,” Analyses of Social Issues and Public Policy, vol. 3, no. 1, pp. 65–86, 2003.
[6] F. Bardhi and G. M. Eckhardt, “Access-based consumption: The case of car sharing,” Journal of consumer research, vol. 39, no. 4, pp. 881–898, 2012.
[7] S. Shaheen and A. Cohen, “Innovative mobility: Carsharing outlook carsharing market overview, analysis, and trends.” 2020.
[8] T. Litman, “Evaluating carsharing benefits,” Transportation Research
Record, vol. 1702, no. 1, pp. 31–35, 2000.
[9] S. A. Shaheen and A. P. Cohen, “Carsharing and personal vehicle services: worldwide market developments and emerging trends,”
International journal of sustainable transportation, vol. 7, no. 1, pp. 5–
36 | REFERENCES
[10] C. Rydén and E. Morin, “Mobility services for urban sustainability: Environmental assessment,” Moses Report WP6, Trivector Traffic AB,
Stockholm, Sweden, 2005.
[11] G. Seeger and M. Bick, “Mega and consumer trends-towards car-independent mobile applications.” in ICMB, 2013, p. 27.
[12] O. Arazy, O. Nov, and N. Kumar, “Personalityzation: Ui personalization, theoretical grounding in hci and design research,” AIS Transactions on
Human-Computer Interaction, vol. 7, no. 2, pp. 43–69, 2015.
[13] T. Ericsson and M. Nilqvist, “A personalized car: A study on how to apply personalization to a driver environment,” 2006.
[14] J. Kim, S. Kim, and C. Nam, “User resistance to acceptance of in-vehicle infotainment (ivi) systems,” Telecommunications Policy, vol. 40, no. 9, pp. 919–930, 2016.
[15] L. Console, I. Torre, I. Lombardi, S. Gioria, and V. Surano, “Personalized and adaptive services on board a car: an application for tourist information,” Journal of Intelligent Information Systems, vol. 21, no. 3, pp. 249–284, 2003.
[16] D. J. Wheatley, “Beyond the desktop: and into your vehicle,” in CHI’00
Extended Abstracts on Human Factors in Computing Systems, 2000, pp.
43–44.
[17] M. Kuemmerling and C. Heilmann, “Seamless mobility: Individual mobility profiles for a better usability of shared vehicles,” in
International Conference on Human-Computer Interaction. Springer,
2013, pp. 318–322.
[18] J. Sonnenberg, “Service and user interface transfer from nomadic devices to car infotainment systems,” in Proceedings of the 2nd International
Conference on Automotive User Interfaces and Interactive Vehicular Applications, 2010, pp. 162–165.
[19] Carplay. [Online]. Available:https://www.apple.com/ios/carplay/
[20] Android auto. [Online]. Available: https://www.android.com/auto/
REFERENCES | 37
[22] accenture, “Mobility as a service,” 2018. [Online]. Available: https: //www.accenture.com/us-en/insight-mobility-automotive-ecosystem
[23] Volvo, “Volvo car group.” [Online]. Available: https://group.volvocars. com/company
[24] T. McCurdie, S. Taneva, M. Casselman, M. Yeung, C. McDaniel, W. Ho, and J. Cafazzo, “mhealth consumer apps: the case for user-centered design,” Biomedical instrumentation & technology, vol. 46, no. s2, pp. 49–56, 2012.
[25] R. E. Boyatzis, Transforming qualitative information: Thematic analysis
and code development. sage, 1998.
[26] K. Anderson, O. Burford, and L. Emmerton, “Mobile health apps to facilitate self-care: a qualitative study of user experiences,” PloS one, vol. 11, no. 5, p. e0156164, 2016.
[27] F. Ciari, B. Bock, and M. Balmer, “Modeling station-based and free-floating carsharing demand: Test case study for berlin,” Transportation
Research Record: Journal of the Transportation Research Board, vol.
2416, pp. 37–47, 12 2014. doi: 10.3141/2416-05
[28] M. Nourinejad and M. J. Roorda, “Carsharing operations policies: a comparison between one-way and two-way systems,” Transportation, vol. 42, no. 3, pp. 497–518, 2015.
[29] S. A. Shaheen, N. D. Chan, and H. Micheaux, “One-way carsharing’s evolution and operator perspectives from the americas,” Transportation, vol. 42, no. 3, pp. 519–536, 2015.
[30] S. Harms and B. Truffer, “The emergence of a nation-wide carsharing co-operative in switzerland,” A case-study for the EC-supported rsearch
project “Strategic Niche Management as a tool for transition to a sustainable transport system”, EAWAG: Zürich, 1998.
[31] H. Becker, F. Ciari, and K. W. Axhausen, “Comparing car-sharing schemes in switzerland: User groups and usage patterns,” Transportation
Research Part A-policy and Practice, vol. 97, pp. 17–29, 2017.
38 | REFERENCES
[33] Frost&Sullivan, “Future of carsharing market to 2025,” 2016. [Online]. Available: http://www.frost.com/sublib/display-report.do? id=MB4D-01-00-00-00
[34] Share now. [Online]. Available: https://www.share-now.com
[35] J. E. Burkhardt and A. Millard-Ball, “Who is attracted to carsharing?”
Transportation Research Record, vol. 1986, no. 1, pp. 98–105, 2006.
[36] M. Prieto, G. Baltas, and V. Stan, “Car sharing adoption intention in urban areas: what are the key sociodemographic drivers?”
Transportation Research Part A: Policy and Practice, vol. 101, pp. 218–
227, 2017.
[37] J. Heikkinen, E. Mäkinen, J. Lylykangas, T. Pakkanen, K. Väänänen-Vainio-Mattila, and R. Raisamo, “Mobile devices as infotainment user interfaces in the car: contextual study and design implications,” in
Proceedings of the 15th international conference on Human-computer interaction with mobile devices and services, 2013, pp. 137–146.
[38] K. Udovicic, N. Jovanovic, and M. Z. Bjelica, “In-vehicle infotainment system for android os: User experience challenges and a proposal,” in 2015 IEEE 5th International Conference on Consumer
Electronics-Berlin (ICCE-Electronics-Berlin). IEEE, 2015, pp. 150–152.
[39] M. Chalmers and I. MacColl, “Seamful and seamless design in ubiquitous computing,” in Workshop at the crossroads: The interaction
of HCI and systems issues in UbiComp, vol. 8, 2003.
[40] T. Kujala, “Browsing the information highway while driving: three in-vehicle touch screen scrolling methods and driver distraction,” Personal
and ubiquitous computing, vol. 17, no. 5, pp. 815–823, 2013.
[41] S. Graf, W. Spiessl, A. Schmidt, A. Winter, and G. Rigoll, “In-car interaction using search-based user interfaces,” in Proceedings of the
SIGCHI Conference on Human Factors in Computing Systems, 2008, pp.
1685–1688.
[42] J. Kim, J. H. Ryu, and T. M. Han, “Multimodal interface based on novel hmi ui/ux for in-vehicle infotainment system,” Etri Journal, vol. 37, no. 4, pp. 793–803, 2015.
REFERENCES | 39
[44] N. Hataoka, M. Araki, T. Matsuda, M. Takahashi, R. Ohtaki, and Y. Obuchi, “Evaluation of interface and in-car speech-many undesirable utterances and sever noisy speech on car navigation application,” in 2008
IEEE 10th Workshop on Multimedia Signal Processing. IEEE, 2008,
pp. 956–959.
[45] D.-h. Kim and H. Lee, “Effects of user experience on user resistance to change to the voice user interface of an in-vehicle infotainment system: Implications for platform and standards competition,” International
Journal of Information Management, vol. 36, no. 4, pp. 653–667, 2016.
[46] M. Schneider-Hufschmidt, U. Malinowski, and T. Kuhme, Adaptive user
interfaces: Principles and practice. Elsevier Science Inc., 1993.
[47] J. Blom, “Personalization: a taxonomy,” in CHI’00 extended abstracts
on Human factors in computing systems, 2000, pp. 313–314.
[48] M. J. Culnan and R. J. Bies, “Consumer privacy: Balancing economic and justice considerations,” Journal of social issues, vol. 59, no. 2, pp. 323–342, 2003.
[49] J. Grossklags and A. Acquisti, “When 25 cents is too much: An experiment on willingness-to-sell and willingness-to-protect personal information.” in WEIS, 2007.
[50] J. Sutanto, E. Palme, C.-H. Tan, and C. W. Phang, “Addressing the personalization-privacy paradox: an empirical assessment from a field experiment on smartphone users,” MIS quarterly, pp. 1141–1164, 2013. [51] M. Hasenjäger and H. Wersing, “Personalization in advanced driver assistance systems and autonomous vehicles: A review,” in 2017 ieee
20th international conference on intelligent transportation systems (itsc).
IEEE, 2017, pp. 1–7.
[52] J. Letchner, J. Krumm, and E. Horvitz, “Trip router with individualized preferences (trip): Incorporating personalization into route planning,” in
AAAI, 2006, pp. 1795–1800.
40 | REFERENCES
[54] G. D. Abowd and E. D. Mynatt, “Charting past, present, and future research in ubiquitous computing,” ACM Transactions on
Computer-Human Interaction (TOCHI), vol. 7, no. 1, pp. 29–58, 2000.
[55] J. I. Árnason, J. Jepsen, A. Koudal, M. R. Schmidt, and S. Serafin, “Volvo intelligent news: A context aware multi modal proactive recommender system for in-vehicle use,” Pervasive and Mobile Computing, vol. 14, pp. 95–111, 2014.
[56] S. R. Garzon, “Intelligent in-car-infotainment system: A prototypical implementation,” in 2012 Eighth International Conference on Intelligent
Environments. IEEE, 2012, pp. 371–374.
[57] F. D. Davis, “A technology acceptance model for empirically testing new end-user information systems: Theory and results,” Ph.D. dissertation, Massachusetts Institute of Technology, 1985.
[58] ——, “Perceived usefulness, perceived ease of use, and user acceptance of information technology,” MIS quarterly, pp. 319–340, 1989.
[59] P. Legris, J. Ingham, and P. Collerette, “Why do people use information technology? a critical review of the technology acceptance model,”
Information & management, vol. 40, no. 3, pp. 191–204, 2003.
[60] V. Krishnaraju, S. K. Mathew, and V. Sugumaran, “Web personalization for user acceptance of technology: An empirical investigation of e-government services,” Information Systems Frontiers, vol. 18, no. 3, pp. 579–595, 2016.
[61] X. Guo, X. Zhang, and Y. Sun, “The privacy–personalization paradox in mhealth services acceptance of different age groups,” Electronic
Commerce Research and Applications, vol. 16, pp. 55–65, 2016.
[62] E. Park and K. J. Kim, “Driver acceptance of car navigation systems: integration of locational accuracy, processing speed, and service and display quality with technology acceptance model,” Personal and
ubiquitous computing, vol. 18, no. 3, pp. 503–513, 2014.
[63] W. Geldmacher, V. Just, J. Kopia, and A. Kompalla, “Development of a modified technology acceptance model for an innovative car sharing concept with self-driving cars,” in BASIQ International Conference.
New Trends in Sustainable Business and Consumption-2017, edited by Rodica Pamfilie, Vasile Dinu, Laurent,iu Tăchiciu, Doru Ples,ea, Cristinel
REFERENCES | 41
[64] S. Osswald, D. Wurhofer, S. Trösterer, E. Beck, and M. Tscheligi, “Predicting information technology usage in the car: towards a car technology acceptance model,” in Proceedings of the 4th International
Conference on Automotive User Interfaces and Interactive Vehicular Applications, 2012, pp. 51–58.
[65] Y. Lee, K. A. Kozar, and K. R. Larsen, “The technology acceptance model: Past, present, and future,” Communications of the Association
for information systems, vol. 12, no. 1, p. 50, 2003.
[66] J. Spradley, “The ethnographic interview holt reinhart and winston,” New
York, 1979.
[67] V. Braun and V. Clarke, “Using thematic analysis in psychology,”
Qualitative research in psychology, vol. 3, no. 2, pp. 77–101, 2006.
[68] M. Diehl and W. Stroebe, “Productivity loss in brainstorming groups: Toward the solution of a riddle.” Journal of personality and social
psychology, vol. 53, no. 3, p. 497, 1987.
[69] Official page of framer. [Online]. Available: https://framer.com
[70] K. Renaud and J. Van Biljon, “Predicting technology acceptance and adoption by the elderly: a qualitative study,” in Proceedings of the 2008
annual research conference of the South African Institute of Computer Scientists and Information Technologists on IT research in developing countries: riding the wave of technology, 2008, pp. 210–219.
[71] J. Nielsen, “Guerrilla hci: Using discount usability engineering to penetrate the intimidation barrier,” Cost-justifying usability, pp. 245– 272, 1994.
[72] B. Friedman, P. H. Kahn, J. Hagman, R. L. Severson, and B. Gill, The
Watcher and the Watched: Social Judgments About Privacy in a Public Place. Springer London, 2009.
Appendix A: Appendix | 43
Appendix A
Appendix
Question Type
Could you describe the steps of renting a car? Task-related
What did you do after getting in the car? Task-related
What is your experience of connecting the smartphone to the car Experience What do you usually do with the phone connected to the car? Typical
Why do you use this app in the car? Use
Table A.1: Interview questions Sequence Task
1 Get on the car and connect the phone 2 It is rainy. Find the wiper instruction
3 Receive Alice’s location
4 Tell Alice you will pick up her in 10 minutes 5 Set the mountain as the destination
