2.1. Theoretical Background
The world is evolving to a hyper-connected society. Transportation as one of the main society sectors is adopting innovative changes in the existing processes and services. Internet of Things (IoT) and location-based big data facilitated transactions that were not possible in the past. In the closed-innovation concept [28
], “technology policy has emphasized the supply side of development, but in the open innovation paradigm on the contrary, it must work on the demand side. The concept of ‘demand articulation’ was effective in formulating corporate policies for technology and market development, and also in government policies for accelerating the commercialization process of emerging technologies” ([10
], p. 3). Attempts to create a dynamic economy are collectively called the Fourth Industrial Revolution [10
The Fourth Industrial Revolution that has been progressing since the start of the 21st century, and features the creative connection between technology and the market in all industries based on IT, that is, the creative and open combination of technology and the market through open innovation, or growth based on the open business model [29
]. Examples include smart cities based on an open innovation or open business model platform, and AV systems that spread throughout the nation from a specific region [30
]. For a converted new market industry, such as the AV industry, a new business model itself leads to market growth, and technology development is required to implement the new business model in a sophisticated manner. In the current emerging period, the effect of the business model is even greater than the effect of technology [31
Research policies in both the public and private sectors are important to guide technological developments under limited resources. Patent development, knowledge transfer, and the commercialization of technology applications are also critical aspects of the development of technology [30
Former studies that concentrated on AVs exemplified how socioeconomic characteristics are correlated with AV acceptance [24
]. Furthermore, scholars investigated the significance of public perceptions, personal characteristics, and attitudes in determining the intention to adopt AVs (e.g., [23
]. Public perception is “simply the type of information obtained from a public opinion survey” ([36
], p.40). Surveys have been done on the public perceptions of AVs, mostly considering people as potential users (e.g., [37
]). Individual attitudes toward AVs are a key contributor of the technology demand, governing policies and future investments in infrastructure [41
Some scientists stress the importance of making a distinction between “acceptability” and “acceptance”. According to Jamson [42
], acceptability is “how much a system is liked”, while acceptance is “how much it would be used”. Despite the manifestation of different views toward acceptance and acceptability, they both are recognized to be based on the individuals’ judgement of the system [43
]. Previous research efforts borrowed from the theories and models of user adoption that were developed to clarify the relationships between attitudes and human behavior [17
These models are comprised of the “technology acceptance model” (TAM) [45
], “theory of planned behavior” (TPB) [47
], and the “unified theory of acceptance and use of technology” (UTAUT) [48
]. In some cases, one of these models as a main theoretical frame of reference was extended with another theoretical model as the original theory was not able to signify the distinct properties of AVs [50
]. According to Jing et al. [52
], “the application of theory-based models can provide stronger predictors and explanations about the determinants of AV adoption intention” ([53
], p. 2). Table 1
summarizes the paradigms, contributions, and representative constructs of these theories with their definitions, with reference to some relevant studies.
Additionally, for the sake of examining the adoption likelihood of a novelty by the public, the “diffusion of innovation theory” (DOI) [79
] envisages individuals’ perceptions of the novelty—such as AV [23
]. In this domain, “attitudes” are regarded as psychological tendencies that signify the degree to which individuals like or dislike a particular entity [80
]. Attitude, considered as a moderator of beliefs and intentions, exists in both the theories of TPB and TAM [81
]. Furthermore, Rogers [79
] pointed out that potential users’ overall attitude towards an innovation determines their adoption decisions, and significantly impacts their acceptance of the AVs, which has been acknowledged by the AV acceptance studies as well [82
In this context, “intention to use” is a significant notion as fully AVs are not yet publicly available. Intention to use is formed according to acceptance levels unless the technology becomes tangible and riders gain a chance to experience it, in reality, and are able to make decisions and give consistent and reliable answers to questions concerning actual use [54
]. This method was deliberated upon the interpretation was made that the general public are mostly unfamiliar with AVs, thus potential misconceptions and uncertainties regarding such innovations could affect the accuracy and consistency of reported findings [50
]. This issue points to an inconsistency in the literature regarding predictors of public acceptance and adoption preferences of AVs.
2.2. Research Design
The study undertakes a systematic review of the AV literature to address the main research question of: What are the individual predictors of public acceptance and intention to adopt AVs? The rationale is that although there are some review studies published on this topic [23
], there is not any comprehensive review of the most recent studies in the field. Considering the fact that more than 50% of publications (n = 42; 52.2%) on this topic were published during 2019, covering this gap via providing a comprehensive review of key findings of the latest investigations on AV adoption intention, and developing a conceptual framework, is necessary. This—including the large amount of recently published work in the analysis—differentiates our research from the prior studies. The adopted methodologic approach is a three-phase procedure that is used by many studies [86
In an initial planning phase, we developed the research aim, question, and keywords, and we set the inclusion and exclusion criteria. The research aim was framed to identify and classify individual determinants of AV public acceptance and intention to use and the most common measurement items had been used for each factor. The inclusion criteria target English language peer-reviewed journal articles, available online in full-text and related to the research aim. Studies focusing only on highly and fully automated passenger vehicles with SAE [27
] automation levels of 4 and 5 were included in the review. The exclusion criteria were intended to be publications except for those pointed out in the inclusion criteria. An online search was done through a university’s search engine, enabling access to 393 different databases, comprising Scopus, Science Direct, Web of Science, Wiley Online Library, Transport Research International Documentation (TRID).
The second phase involves reviewing relevant articles. The initial publication date in the search was not determined, as the earliest surveys on AVs were from 2012 [24
], whereas the end date (when the search was carried out) is December 2019. The query string applied for database searches was: (automated OR autonomous OR driverless OR self-driving) AND (car OR vehicle) AND (perception OR acceptance OR adoption OR intention OR preference OR choice experiment). The keywords focused on the title and abstract of the articles explored. The search firstly resulted in overall 1966 records.
The filtering or screening process for selecting relevant literature pieces (via exclusion and inclusion approaches) forms an integral part of a systematic review. This is done in two stages: (a) Screening, and; (b) Eligibility (Figure 1
In the screening stage, the identified articles were screened and lessened to 579 by applying the primary inclusion and exclusion criteria. This amount was “eye-balled for consistency and accuracy of the keyword search” [88
], and the duplicates were also removed. Next, the abstracts of the remaining 536 articles were evaluated by applying the secondary inclusion and exclusion criteria and the records against the research aim were excluded. Then, in the eligibility stage, the full-texts of 144 articles were retrieved and screened to definitely decide whether the study fits the eligibility criteria of our review. Finally, after exclusion of the irrelevant papers regarding the aim of the study, the number of articles was narrowed down to 80 for the review, categorization, and analysis round.
The current study is founded on a descriptive rather than statistical analysis of literature. As for Yigitcanlar et al. ([87
], p.353), “qualitative techniques of pattern matching and explanation building have been adopted to descriptively categories the journal articles under specific categories” [84
]. In this regard, according to Yin, [84
], pattern matching refers to scanning for commonalities and disparities in which an eye-balling technique is sufficiently convincing to draw a conclusion or categorization. The selection criteria to formulate the categories is itemized in Table 2
. The categorization was adjusted and allocated into three different clusters that may potentially impact the public acceptance and adoption intention of AVs—i.e., demographic, psychological, and mobility behavior characteristics. These individual determinant categories are described in detail in the following section and also their relevant studies were illustrated in Table 3
In the third phase, which involves reporting and dissemination, the findings were critically documented and presented in the form of a review paper. Additional publications on the subject, which were retrieved through conducting forward and backward citation chasing, were simultaneously included as supporting literature to comprehensively elaborate the theoretical background and overall findings of the research. Even if these publications do not meet the selection criteria, this strategy is a invaluable method to identify relevant sources as we are tapping into the expertise and prior research of the authors. Hence, we are essentially browsing a curated list of sources.