A survey on obstacles to the widespread use of connected and automated vehicles

Abstract

Connected and Automated Vehicles (CAVs) are rapidly evolving technology with great benefits such as reducing gas emissions and decreasing traffic congestion. They have the potential to change the traditional transportation industry due to their benefits. However, the implementation phase for CAVs decelerates with the uncertainties of legislation on privacy-preserving and public concerns. Perception of people needs to be understood beforehand. Main concern points like possible attacks and mitigation techniques, and privacy protection should be addressed. Certain regulation system should be implemented, and transportation habits should be considered. After thinking over those points, adaption of CAVs can be achieved more smoothly. In this survey paper, we aim to shed light on the obstacles to the widespread use of CAVs by collecting existing literature and creating a sophisticated bouquet of the issues. Public perception, common attacks and mitigation techniques, privacy protection, regulations, and possible transportation habit shifts related to CAVs are examined. With the information gathered from this survey, manufacturers and policymakers can determine an influential pathway for the development of CAVs.

Keywords

CAV privacy cyber security

1. Introduction

Existing transportation habits result with ineffectual traffic flow. Poorly planned road trips and inexperienced drivers are some of the reasons behind that. Unnecessary time is spent on traveling. This also increases fuel consumption and environmental pollution. Another problem is elders and the ones who are unable to drive alone depend on a driver for traveling. All these problems raise a need for a shift in the transportation industry. New transportation habits should be developed to solve the existing problems. Connected and Automated Vehicles (CAVs) are the new era in transportation industry. CAVs have various benefits such as reducing traffic congestion, easing the travel of nondrivers (elders and children etc.), diminishing fuel consumption, and most importantly scaling down the traffic accidents as the majority of the current accidents are caused by human errors [20,58]. These benefits lay a burden on CAVs to shape the next transportation habits. However, a new technology comes with new worries. The uncertainties of CAVs make people doubt using them. It is essential to solve obstacles to the widespread use of CAVs so that transformation to this technology can be achieved smoothly.

Fig. 1.

SAE levels of driving automation.

The name of CAV has two parts i) connected and ii) automated. Connected means there exists communication between vehicles and infrastructures to share data about the environment that will help the decision-making process. Automated means providing a traveling experience without the need for a human driver. Society of Automotive Engineers (SAE) defines 6 levels of automation where level 0 is no automation and level 5 is fully automated vehicles [52].

Vehicles with the first three SAE levels (0, 1, 2) are commonly seen on the road. According to World Economic Forum [6] the percentage for vehicles corresponding to Levels 0, 1 and 2 are 14.4, 55.7 and 29.9, respectively. By 2025, the same study expects the percentages as 1.6, 63.6 and 34.1.

As it can be seen from Figure 1 [52], the former three levels require constant driver control, while the letter three introduce automated driving features. Even at Level 3, the presence of a driver is still required. Poorly understanding the levels of autonomy and misaddressing what the levels actually correspond to in the advertisements of the vehicles result in embezzlement. To give an example on the road, Tesla’s Autopilot, advanced driver assistance technology suite, corresponds to Level 2 in this scale while Mercedes-Benz is world’s first automotive company to certify Level 3 system, with Drive Pilot, for U.S. market [14] [35]. In February 2023, Musk, the CEO of Tesla, faced a lawsuit from stakeholders complaining about him overstating the capabilities of Autopilot, which resulted in Tesla’s share price falling [54]. Even though, the name “Autopilot” seems to suggest a driverless experience, Level 2 vehicles still require a human assistance and taking hands off the wheel is illegal. The focal point of the discussions within the paper pertains predominantly to fully-automated vehicles, specifically those categorized as Level 4 and Level 5 autonomy. It is noteworthy that, unless explicitly stated otherwise, the terms “CAVs” in this context exclusively refer to Level 4 and Level 5 vehicles. This deliberate focus is attributed to the inherent technical challenges that persist in these vehicular systems, necessitating thorough examination within the existing literature. The heightened complexity and unresolved issues inherent to Level 4 and Level 5 automation render them particularly salient subjects of investigation and analysis in this discourse.

There are assorted reasons hampering the widespread use of CAVs. The first and foremost reason is the prejudice of people. People tend to feel nervous about the possibilities and the outcomes of new technology. Uncertainties about how the vehicle behaves under unexpected circumstances affect their opinion negatively on adapting CAVs into their lives. Also, the rise in the importance of personal data, especially during pandemic, and the results of data leakage make them feel hesitant [23]. Understanding the public’s opinion and shaping the development phases accordingly benefits manufacturers and vendors for providing a better service. The paper highly emphasizes the importance of public perception and brings together the existing literature to have a wider perspective about it.

Cyber attacks constitute the major safety concerns of CAVs. Cyber attacks might cause malfunction of a vehicle that results with disruption of the traffic flow or, even worse, a traffic accident that may cause a fatal result. Control mechanisms of a vehicle, such as a brake pedal or lock system, might be controlled remotely by an adversary. This paper brings together the common cyber attacks that target CAVs and possible mitigation techniques.

Collected data could also become the target since it constitutes a valuable information source for malicious people. Privacy protection of the collected and processed data should be considered beforehand. Methods to provide the privacy of the data is included in this paper.

Legislation phases of privacy protection and limits on data sharing should be determined and regulated at the early stages of the implementation. Regulation for CAVs began in some countries. However, the nature of data flow requires globally applied comprehensive legislation. Regulations are commented in this paper.

To predict the possible shift in the transportation industry, insight of the existing habits and foresight for the change in the transportation industry is also mentioned in this paper.

Aforementioned topics cover the obstacles to the widespread use of CAVs together with suggestions to overcome the drawbacks step by step. The aim of this study is to define the prominent factors of drawbacks on the extensive use of CAVs. This paper brings out the existing impediments to the reader. The findings are classified in order to ease the reading process. The contribution of this study is to collect the existing literature and create a sophisticated bouquet of the obstacles to the widespread use of CAVs. To do so, the findings are divided into five sections.

Public Perception (§3) as the acceptance of a new technology strongly depends on public perception, this section details the related academic work and findings to display the current situation.

Common Attacks and Mitigation Techniques (§4) elaborates on the literature on the most common cyber attacks and the possible threats that CAVs face together with recent solutions to existing communication problems.

Privacy (§5) focuses on data protection and privacy preserving methods for CAVs.

Regulations (§6) recommendations for regulations are detailed.

Transportation Habits (§7) a probable shift in transportation habits is detailed.

Table 1 shows the examined literature findings organized under defined titles. These findings contain the obstacles on the widespread use of CAVs. Some of them include possible solutions. Details of the findings are mentioned in the respective sections.

Table 1

List of reviewed studies grouped under the related subtopics.

Topic	Subtopic	Studies	Content	Findings
Public perception	a) Perception model	[13,59]	Details the acceptance and public perception of a new technology	Lack of knowledge cause dread, education is needed to overcome it
	b) Surveys	[3,9,22,27,38,44,47–50,61]	Surveys that are conducted to collect opinion of people about CAVs
	c) Suggestions	[31,32,34]	Suggestions to the manufacturers and policymakers about CAVs with the sight of public perception
Common attacks and mitigation techniques	a) Experiments	[7,25,42,46,53,63,65]	Experiments that perform the examined attack on a vehicle or via a simulation	It is an ongoing process, ML solutions are on peak recently
	b) Thematic research	[21,39,41,43,56]	Existing cyber attacks and possible solutions to them
	c) Technical improvements	[30,66]	Proposes a technical improvement to make CAVs more efficient
Privacy		[1,8,18,36,62]	CAVs process huge amounts of data in which vast majority of it contains sensitive information these studies concern this topic	Privacy preserving methods like encryption and anonymization should be adapted in the early stages
Regulations		[2,10,12,16,26,28]	CAVs require some regulations regarding to use in daily life, existing regulations are mentioned and compared in these studies	Global regulations are needed to match with the nature of data flow
Transportation habit		[29,33,55,60]	CAVs are believed to be changing the old transportation habits, these studies investigate that topic	Demographic and physical changes should be considered beforehand

2. Methodology

This thematic survey employs a systematic and comprehensive approach to identify, analyze, and synthesize obstacles to the widespread use of CAVs across a diverse range of sources. The research framework is originated by Braun and Clarke’s Thematic Analysis [5] as their method of finding and classifying data for a survey aligns with our criteria.

2.1. Search strategy

A systematic search strategy is employed to identify relevant literature. Databases such as IEEE Xplore, ScienceDirect and SpringerLink are searched using key terms: “Adoption of CAVs”, “Challenges on CAVs”, and “Current developments on CAVs”. The use of up-to-date resources is prioritized to reflect the most relevant aspects of the topic. After collecting resources, we carefully handpicked the ones to be used in our paper. Additionally, we employed the snowball method to broaden the scope of identifying suitable articles, wherein the references of selected articles were examined to discover relevant literature.

The primary data for this survey consist of textual information from selected sources. Data collection involves the extraction of relevant passages, quotes, or excerpts related to the identified theme. Each source is carefully documented, including publication details, to maintain transparency and traceability.

2.2. Writing process

After identifying the resources, they are categorized into subsections based on relevant subjects. To emphasize each point adequately, five subjects are identified, each containing different aspects of the topic. The paper is then structured around these, in the following sections.

3. Public perception

Public perception is believed to be shaping the acceptance and development phases of a new technology. Adoption of new technology requires the consent of people before taking a position in daily life. Public demand will shape the acceleration of CAV development. Informing people about the possibilities of CAV is the first step toward shaping their perspective.

3.1. Perception model

Davis [13] develops and validates new scales for two specific variables, perceived usefulness and perceived ease of use, which are suggested to be fundamental determinants of user acceptance. Davis’ study was conducted back in 1989 to examine the acceptance of computers. Later, it became the backbone for acceptance of new technologies and still preserves its validity. His study results show that perceived usefulness and ease of use are the top criteria for acceptance of new technology. If people are informed about the benefits of CAVs, then it would ease the process of accepting it.

In another study [59], authors review user acceptance literature and discuss eight prominent models, then empirically compare them and their extensions, thus formulate a unified model. Their study examines eight models that are used dominantly for understanding public acceptance of new technologies. Then, the authors create a new model using the former eight models. They test all of these models, using data from four different organizations. The results show that the unified model outperforms the existing eight models. That means the unified model performs a more accurate behavior model for user acceptance of new technologies and helps to understand the public view. Their model is utilized in many other studies about public perception of CAVs. Such a unified model is the key to understand public’s acceptance in order to adapt CAVs into daily life.

3.2. Perception surveys

Various surveys collect the public perception about CAVs. For the sake of collecting meaningful feedback, respondents generally picked from places where people are already familiar with the CAV concept or were informed about the technology prior to the survey.

Schoettle and Sivak [49] document the results of a study that examines the public opinion about self-driving vehicles in China, India, and Japan. The survey yielded responses from 610 respondents from China, 527 respondents from India, and 585 respondents from Japan. The respondents were generally aware of the technology and had positive initial thoughts about it. Results show that people generally feel positive toward CAVs. However, they hesitate possible safety issues, especially they consider the vehicles that lack drivers to be unsafe due to system errors. People would like to use this technology in their vehicles in case they are not paying extra amounts for this upgrade. This study gives an idea to manufacturers about the market point of CAVs and people’s perception of them.

Kyriakidis et al. [27] investigate user acceptance, concerns, and willingness to buy partially, highly, and fully automated vehicles through a 63-question internet-based survey, that collects 5000 responses from 109 countries, giving a global perspective on the perception of CAVs. Results of this survey show that people fear the cyber security of CAVs and the possible outcomes of being hacked. Also, results show that citizens of more developed countries doubt about data sharing among CAVs due to possible leakage. With a large number of respondents from different countries, this study sheds the light on the global perception of CAVs.

Bansal et al. [3] gathered the findings of an internet-based survey that polled 347 Austinites to understand their opinions on smart-car technologies and strategies. The survey collects the opinions of Austinites on accepting CAVs and their willingness to pay (WTP) for owning a CAV. Results of this survey show that respondents perceive the main benefit of CAVs to be reducing accidents. Yet, they are mainly concerned about the malfunction of the vehicles, especially with higher levels of automation. The survey results can be used for planning the market point of CAVs and strategically adapting to the transportation system.

Piao et al. [44] examines public perception of the use of automated vehicles in urban areas. This study is a part of the ongoing evaluation activity on CAVs in urban areas. People had positive feelings about automated shuttles around the city due to their cost efficiency. Even though some safety concerns about CAVs were present, like traveling at nighttime, results show that the general perception about the use of CAVs was positive. Introducing the CAVs in some urban areas to see the public perception is important before fully launching them on the roads. Results of public perception can help the improvement of the adaptation process.

Surveys from different parts of the world help us to better understand the general perception of the CAVs. Common ground to those surveys is people are mostly familiar with this technology. Results of the surveys show that familiarity increases the acceptance factor, but even though people like the concept they fear on malfunction of the vehicles especially the ones that does not allow interference of the driver.

There are also some surveys that provide information about the CAVs to the respondents prior to conducting the survey. Nordhoff et al. [38] present the results of an interview study with 30 users who took a tour with an automated shuttle on the EUREF (Europaisches Energieforum) campus in Berlin- Schoneberg to obtain an in-depth understanding of the acceptance of automated shuttles for public transport systems. Respondents were generally positive about using AVs. Although current vehicles do not satisfy their expectations, people expect AVs to take place on the roads in the short term. Experiencing the ride with autonomous vehicles allows people to have positive attitudes toward the technology and helps them to overcome the dread.

Salonen and Haavisto [48] applied Harry Triandis’ Theory of Interpersonal Behavior (TIB) to interviewees whom the passengers who traveled a predefined route in a driverless shuttle bus. TIB acknowledges human behavior as habitual responses to situations and intentions rather than always being rational. Thus, this study examines the collected data with that in consideration. Results show that when respondents traveled with an autonomous vehicle, they were surprised by its safety level and felt positive about using driverless vehicles. These results show the importance of familiarity with the technology in the acceptance process.

Chikaraishi et al. [9] expands a study on existing literature by directly examining which aspects of Automated Vehicle (AV) use and function most affect risk perception. For that purpose, short animated video clips of AVs are shown to participants. Then, they were surveyed about their attitudes towards AVs. Three potential risk factors are animated in the videos: system error, hacking and unexpected events. Results of this survey show that even though the perceived advantages of the AVs generally overcome the perceived risks, there is still a respectable amount of perceived risk which is mainly led by the dread of the unfamiliarity of new technology. People should be informed about the technology before adapting it to daily life to overcome the dread factor.

Hilgarter and Granig [22] aim to explore the transportation habits of current passengers, the perception of the public on how AVs can be used in daily life, and the general attitude of the public towards it, by examining the perceived safety of passengers after riding an autonomous shuttle with a maximum SAE level of 3 in the setting of mixed traffic. This qualitative study is conducted with 19 participants varying in age to examine their perspectives on AVs. To do so, participants were given a tour in real traffic with an autonomous shuttle. Then surveyed about their perceptions of this experience. Respondents had positive feelings about the safety level of the vehicle after being given the tour. This shows that familiarity with CAVs positively effects the attitude toward it. According to the findings, both people who had previous experience with the CAVs and people who were recently introduced to CAVs are affected positively.

It is not only the drivers that create the current traffic on the roads. There are also non-drivers like pedestrians and bicycle-riders. Their perception of CAVs is also important for shaping CAVs’ future. Pyrialakou et al. [47] point out the lack of literature on a detailed exploration of the safety perceptions of road users who will interact with AVs, and conduct a survey in Phoenix, Arizona, to address this gap. This survey was conducted with the locals who are already familiar with AVs by seeing them on the road in their daily lives. Findings of this survey show that an individual’s level of the perceived safety of AVs is linked to the level of awareness. Also, the automation level of the vehicle affects the perceived safety. Familiarity with new technology could be provided by seeing it around or getting educated about it. This helps people to understand and accept CAVs.

CAVs positive effect on the environment can also be a driver for people who care about it. Wu et al. [61] aim to understand consumer attitudes toward autonomous, connected, and electric vehicles (ACEVs), using data collected through a survey conducted in China. Results of this survey show that people were finding positive effects of ACEVs on being eco-friendly, reducing driving fatigue, and allowing non-drivers to travel more. The benefits of ACEVs let people have positive feelings toward them. However, safety concerns and costs make them doubt having one. This study shows the general perception of ACEVs in China, where people are familiar with the technology. The general outcome of these studies shows that people are positive about the benefits of CAVs and believe it to be changing existing transportation habits. Their common responses suggest that knowledge about CAVs affects the perception positively.

Media plays an effective role in the acceptance phase. People can easily be influenced by the media when they are uninformed about technology [50]. Posting about the malfunction of a vehicle and the accidents that occurred might attract public attention. However, it might hurt the safety perception of CAVs. The media should inform people about the reasons for the accident and the development phases that can solve those problems. People also should be educated about the fundamentals of CAVs.

3.3. Suggestions

Some surveys use the collected data to provide a solution to the existing problems. In this case, Liu et al. [32] employ 36 semi-structured elite interviews to explore the diverse dimensions of user acceptance through the lens of the well-informed CAV experts that already anticipate problems and look for their solutions. This study divides the acceptance of CAVs into six sub-topics: awareness, user and vendor education, safety, responsibility, legislation, and trust. Then, conducts interviews with people who are either working in the field or academicians who work in this area. The results of this survey mainly emphasize the importance of education. They suggest that education should be held mandatory for both vendors and customers. This result supports the findings of the aforementioned studies on the importance of knowledge in technology. Respondents suggest that education about CAVs could be held with a program like something that is covered in the driver’s license program.

In another study [31], Liu aims to identify and create an in-depth understanding of the cyber security and privacy issues associated with CAVs by employing a sequential mixed method approach, with a qualitative phase followed by a survey-based phase looking to model the factors underpinning CAV acceptance. A theory-based extended technology acceptance 5 model (TAM) model was developed for this online survey. Based on the findings, policy recommendations are provided together with market-entry recommendations for CAVs. This survey could be a helpful resource for policymakers as well as manufacturers.

Maeng et al. [34] analyze the types of information security threats that the consumers consider the most dangerous and consumer preferences for the information security solutions that protect their CAVs. This survey collects information about public perception of vulnerabilities of CAVs and their WTP for overcoming those flaws. Results of this survey show that consumers perceive protective functions to be the most important, followed by convenience and authentication. Customers’ perception gives manufacturers opinions about where to invest more. The opinion of people is positively affected by their knowledge of the technology. Otherwise, they fear the technology and create the biggest obstacle to the development of CAVs. Mandatory education programs are recommended to overcome the dread. People are also concerned about the malfunction of a vehicle due to cyber attacks.

Main finding of this section is familiarity to the technology positively affect people’s perception to CAVs. Educating people on this topic and giving them the opportunity to experience with this technology can help them to better adapt to it. Also, people mostly worry about the malfunction of the vehicle under a cyber attack and protection of their personal data.

4. Common attacks and mitigation techniques

Attacks that target CAVs might cause misbehavior of the vehicle, disturb traffic flow and lead to accidents. Or might be used for stealing personal data of the user. The motivation behind the attack might be different. However, if the attack leads to success, it might result in a fatal accident. The outcome might cause a sizable amount of financial loss to the manufacturers as well.

4.1. Existing attack attempts

CAVs have various sensors like LiDAR, radar, GPS, and cameras to collect environmental data. All these sensors bring vulnerabilities to the security of the vehicles. There are some studies that perform attacks to those sensors and suggest some prevention mechanisms. Petit et al. [42] present remote attacks on camera-based systems and LiDAR using commodity hardware and propose software and hardware countermeasures that improve sensors’ resilience against those attacks. Sensors in CAVs enable the vehicle to observe its surroundings however they are vulnerable to versatile attacks like blinding, jamming, and spoofing. This research proposes software and hardware precautions against those attacks. Collected data from sensors are used for making decisions about the next move of the vehicle. Therefore, the reliability of sensors affects the safety of vehicles.

In their paper [46], Psiaki and Humphreys describe sets of spoofing attack methods and defense methods in detail, and they develop an attack/defense matrix that documents which defense techniques are effective against the diverse type of attack types, then recommendations are provided to improve the offerings of commercial off-the-shelf receivers from the current situation. Global navigation satellite signals (GNSS) can malfunction with spoofing attacks. Special receivers can provide defense mechanisms to those types of attacks. The paper reviews the spoofing attacks and possible mitigation techniques against them to protect GNSS. GNSS data are widely used in vehicular systems. It is essential to receive correct information to make meaningful decisions. Protecting GNSS against spoofing attacks is necessary for improving the reliability of CAVs.

Zhang et al. [65] present an experimental analysis of the security of vehicles with internet connections and propose an approach for controlling a car through onboard diagnostic (OBD) injection. OBD port is a plug that allows obtaining system information about the vehicle. OBDs are usually directly connected to ECUs of a vehicle which can cause malfunction when accessed. Their study shows the possible vulnerabilities that OBDs have. It is important to consider the vulnerabilities of the system to improve reliability.

Jo et al. [25] indicate the security problems of an Android OS based telematics system by using a device whose firmware is offered on a public website, then they follow up with attack experiments using a real vehicle. Android OS has been introduced to vehicles with its vulnerabilities. Their paper examines remote attacks on an open-sourced device to ease the analysis and then performs the attacks on a real vehicle. CAVs have hundreds of Electronic Control Units (ECUs) and reliable operating systems need to be used to provide safer systems. This study demonstrates that insecurity of a component in the ecosystem might also affect the security of a CAV.

Sitawarin et al. [53] present the unreliability of ML mechanisms on detecting the traffic labels with adversarial attacks. The authors examine the response of ML algorithms when classifying traffic labels and propose an attack method where they provide malfunctioned ads and logos to fool the classification. Their attack method succeeds with the rate of 95% for classifying custom signed labels as traffic labels. This is a huge risk for CAVs as they use traffic labels to make decision about the movement.

Yang et al. [63] take the first attempt to investigate the false data injection (FDI) attacks on a networked radar system. Then, they propose a novel data fusion algorithm to mitigate FDI. Networked radar systems (NRS) are vulnerable to different types of attacks including electronic countermeasure (ECM) jamming and FDI attacks. The majority of the existed research is done on ECM. However, the paper focuses on FDI since it is harder to detect. They first perform an injection attack and then propose an algorithm to combat this attack. NRS are commonly used in CAVs and their reliability should be considered. CAVs also have a network system to communicate both with vehicles (V2V) and other infrastructures (V2I).

Cao et al. [7] perform the first security study of LiDAR based perception in AV settings and find that blindly applying LiDAR spoofing are insufficient to achieve success thus they explore the possibility of strategically controlling the spoofed attack to fool the machine learning model. Regular spoofing attacks are prevented with machine learning-based object detection mechanisms. This study formulates spoofing attacks to trick machine learning mechanisms with a success rate of 75%. Then it also performs a case study to understand the effects of the attack and discusses the improvement methods. LiDAR plays a crucial role in decision-making by providing images of the surroundings. Falsifying the input of LiDAR sensors can lead a vehicle to suddenly brake in ongoing traffic, which might result in fatal accidents. Therefore, it is crucial to have secure sensors.

Table 2
Attacks and mitigation techniques related to CAVs.

Category Types of attacks Mitigation techniques

Sensor attacks Blinding, jamming, spoofing on sensors that CAVs have like LiDAR, radar, GPS, and cameras [7,42,46,63]

Tampering attacks Attacks that can be performed with an access to the vehicle like vehicle control through OBD injection [65]

Software system attacks Security issues based on the software that is used like Android OS-based telematics systems in ECUs [25]

Contextual attacks Adversarial attacks to the outside world like manipulation of road signs [53]

Category	Types of attacks	Mitigation techniques
Sensor attacks	Blinding, jamming, spoofing on sensors that CAVs have like LiDAR, radar, GPS, and cameras	[7,42,46,63]
Tampering attacks	Attacks that can be performed with an access to the vehicle like vehicle control through OBD injection	[65]
Software system attacks	Security issues based on the software that is used like Android OS-based telematics systems in ECUs	[25]
Contextual attacks	Adversarial attacks to the outside world like manipulation of road signs	[53]

Table 2 provides a categorization of attacks specifically targeting CAVs. It is noteworthy that comprehensive classifications of network attacks can be found in esteemed sources such as MITRE [57] which meticulously detail such attacks. Consequently, our paper intentionally excludes discussions on well-documented security issues and instead concentrates on addressing cyber physical system specific aspects that receive comparatively limited attention in the existing literature.

4.2. Thematic research

Some studies follow a thematic approach and piece together the existing attacks that target CAVs together with their possible mitigation techniques that exist in the literature. Petit and Shladover [41] investigate the potential cyber attacks specific to automated vehicles, with their special needs and vulnerabilities. This is a thematic research on the cyber attacks and their occurrence possibilities on CAVs. Some mitigation techniques are also provided. This study can provide necessary information for manufacturers and researchers who would like to study in this area.

In a study conducted by Parkinson et al. [39], a large volume of publicly accessible literature is reviewed and compartmentalized based on the vulnerabilities identified and mitigation techniques developed. The study examines the possible cyber threats that CAVs face. Results show that majority of the vulnerabilities found in CAVs are detected by white-hat hackers. If attacks were performed by malicious people, then the outcomes could cost financial loss to the manufacturers or fatal accidents could occur.

He et al. [21] overview different passive and active cyber security attacks which may be faced by CAVs, also present solutions to each of these attacks based on the current state-of-the-art, then discuss future improvements in the field of CAV cyber security. This study brings together existing computer security issues related to CAVs. Extensive research like this helps create a resource to look up in this area.

Philipsen et al. [43] cover the relevant attacks and threats to modern vehicles and present a security analysis with potential countermeasures. Authors emphasize the responsibility of manufacturers for making safe and reliable CAVs for customers. The paper covers the potential threats and defense mechanisms.

Sun et al. [56] provide a comprehensive survey on the cyber security in the environment of CAVs to highlight security problems and challenges. The paper first classifies the existing cyber security flaws of CAVs and discusses the protection mechanisms. They also contribute by a discussion of the safety standards. It is an inclusive study that can be helpful for both manufacturers and researchers.

Overall, the studies show that white-hat hackers detect most of the known vulnerabilities of CAVs. That leads to improvements on the security flaws. However, malicious people are also trying to take over the CAVs with different motivations like stealing the vehicle, causing a public disturbance, hurting the vehicle owner, and using stored data for malicious purposes. There is an ongoing battle between manufacturers and malicious people. This forces the constant improvement of vehicles and makes them more secure. That will help people to trust and use CAVs in daily life. Users and vendors also should be informed about those attacks to a basic extent to minimize the possibility of human flaws like possible phishing attacks. Gaining the necessary knowledge will not only help the vehicles to be safer but also aid people in feeling the control is in their hands. That will make them feel more confident using a CAV.

4.3. Technical improvements

Technical improvements are required for the safety and reliability of CAVs. Understanding the traffic flow and making decisions is a crucial part of the development. Liu et al. [30] envision that future CAVs necessitate co-design of communication, computation, and control to improve end-to-end performance and they introduce an end-to-end design principle called 4C. With the 4C framework, they aim to provide a programmable unified system for communication, computation, and co-design. The framework is tested under two case studies to prove its usability. This unified model helps to achieve system-level reliability and integration within communication and computation. CAVs use the received data from either another vehicle or infrastructure to determine its current behavior or path planning to provide a better traveling experience. The efficiency and reliability of communication are essential for in-time decision making.

Efficient traffic flow understanding and decision-making capabilities are paramount for CAVs. The utilization of data received from other vehicles or infrastructure plays a pivotal role in determining the vehicle’s behavior and path planning, ultimately contributing to an improved travel experience. To achieve this, the employed algorithms must exhibit speed and security, ensuring fast and reliable decision-making in real-time.

CAVs need to process the received and captured data and make decisions depending on them almost in real-time to comply with the traffic flow. A reason for the utilized algorithms to be fast and secure in making reliable decisions. Zhao et al. [66] propose a minimum delay routing algorithm to minimize the end-to-end packet delay for each vehicular data flow that outperforms the existing routing protocols in quality-of-service performance. The proposed algorithm is tested via a simulation. The results show that it surpasses regular routing protocols in terms of average packet delay. Vehicular communication is needed for CAVs to share data to make decisions about the driving experience.

Timely and secure communication is indispensable for CAVs to prevent accidents and optimize their performance. The need for real-time communication underscores the importance of developing safe and efficient communication protocols. Additionally, as the exchanged data often contains personal information, robust security measures such as encryption and anonymization are imperative to prevent data leakage and uphold user privacy.

In essence, the continued evolution of CAVs relies on the collaborative development of communication, computation, and control systems, as well as the implementation of cutting-edge algorithms and protocols. The pursuit of these advancements is not only crucial for the success of CAV technology but also imperative for ensuring the safety, efficiency, and privacy of future transportation systems. It is noteworthy that the absence of sufficient technical enhancements in these critical domains impedes the adaptation of CAVs.

5. Privacy

CAVs collect and share the personal data of users and people around them. Sensors like LiDAR detect objects around the vehicle. While doing that it also captures the images of the surroundings. For instance the location of the vehicle owner could be determined with possible leakage of these pictures. This information could be analyzed further to obtain the daily habits of the owner and could be used for financial interests, or worse, to cause harm to the owner. Naessens et al. discuss privacy policies and propose tools and mechanisms [36].

Atmaca et al. [1] categorize the emerging privacy challenges and solutions for CAV systems and identify the knowledge gap for future research, which will minimize and mitigate privacy concerns without hampering the efficiency of the functions. The efficiency of CAVs is believed to be increasing with the amount of processed data. However, most of this data contain private information about the owner. The paper brings out the emerging privacy issues related to CAVs.

Xiong et al. [62] propose an edge-assisted privacy-preserving raw data sharing framework to obviate simply encrypting data which introduces a heavy overhead and prevents the risk of data leakage in another vehicle since encrypted data (ciphertext) are then decrypted on another vehicle then the receiver will be fully aware of the sender’s data. The proposed method works as follows. Each image is encrypted into two images and sent to different edge servers. The edge servers employ a revised deep learning model for image processing. Both encrypted features are then sent to the receiver vehicle. Then vehicles produce a meaningful classification result without knowing the original image that is captured by the sender vehicle. Results show that this method can successfully detect objects. Privacy of shared data is sensitive and being able to detect an object without knowing the original image can eliminate the risk of data leakage.

In their study, Girka et al. [18] take the problem of supervised machine learning with deep feedforward neural networks and provide an anonymization algorithm (based on the homeomorphic data space transformation), which guarantees the privacy of the data and allows neural networks to learn successfully. Data owners want to get the most value from what they have. However, they need to obey some privacy protection regulations. That raises the importance of anonymization of the data. The paper proposes an algorithm and shows the results of experiments that have been conducted to see the performance. Anonymization provides a safer platform for data processing in terms of privacy preservation.

Privacy-preserving techniques must be adapted in the early stages of CAV implementation. Recent studies commonly integrate machine learning (ML) methods to solve existing problems. Chellapandi et al. [8] examine the ML methods and their compatibility with CAVs with a focus on Federated Learning (FL). Since ML is a popular area, improvements in that will benefit CAVs by providing better solutions for existing problems.

6. Regulations

Regulations that define the boundaries of CAVs are advised to be determined before seeing the CAVs on the road.

Choi et al. [10] examine the effect of a regulatory policy that requires mandatory disclosure of vulnerabilities by considering a firm that sells software that is subject to potential security breaches and derives the conditions under which a firm would disclose vulnerabilities. Writing flawless software is impractical to achieve. Software needs to be tested and secured even after its launch. Providing an update and disclosing vulnerabilities after fixing a bug only protects the users who install the update. Others are left with the vulnerability and the possible risk of reverse engineering with disclosure. The paper uses a framework to examine the use of mandatory disclosure model effects. Their findings show that disclosure affects customers differently. A regulatory disclosure policy should be adapted in a way to provide safe and reliable software products for CAVs.

Khan et al. [26] quantitatively analyse the Cyber security Regulatory Framework (CRF) for CAVs in an aim to provide a comprehensive resource for policymakers. They model a Stock-and-Flow Model (SFM) with five pillars: the Cyber security Policy Stack, the Hacker’s Capability, Logfiles, CAV adopters, and eSafety Traffic Unit (eSTU) for CAVs-CRF. A visualised aid for CRF is benefical for governments while creating the regulations for CAVs.

D’Amato et al. [12] deals with the behaviour of CAVs under exceptional situations. Main focus area here is the famous “Trolley Car” problem where vehicle needs to decide who will survive in an unavoidable collision. It is expected for vehicle to behave ethically sound and with respect to duty of care owed to other road users even if this conflicts with the defined traffic laws. Authors make suggestions on motion planning algorithms in those extreme scenarios for developers.

Fagnant and Kockelman [16] emphasize the fact that recent regulations of CAVs in the US are held at the state level which may lead to inconsistency in the future. Therefore, the government should create nationally recognized licensing determining appropriate standards for liability, security, and data privacy. US regulations are held differently by each state. However, CAVs do share data among each other and store data via servers that may be found in different states. It is harder to comply with the regulations since vehicles and their data do not stay in one place. The paper suggests US government should adopt nationally applied legislation about CAVs. Data storage and process cover multi-nations. Therefore, legislation should comply with globally.

Lee and Hess [28] review the main safety and liability issues for CAVs with a focus on the rules developed for on-road testing in Australia, the United States, and Germany. As well as they review government policies from Victoria, Australia, and the United States, they provide an appendix on European Union (EU) regulations. The paper reviews the similarities and differences of existing legislation and proposes a harmonized way of applying their best practices. Globally applied regulations provide more effective protection since CAVs require the collection and processing of data all over the world.

Baker et al. [2] compare the emerging regulatory landscape for CAVs in the UK, Germany, and the US, identifying some key areas of commonality and divergence. These three countries are the front runners of the policymaking in the adaption of CAVs. This study compares the regulations in these countries and provides a comparison matrix for ease of use. The paper could be a helpful resource for policymakers in deciding on the legislation for CAVs.

People trust decision makers for regulating the protection of their personal data. Providing privacy-preserving regulations for CAVs in the early implementation phases can ease the adoption process and define the boundaries for manufacturers. People become more willing to integrate CAVs into their lives if data usage is limited by the laws.

7. Transportation habits

CAVs will change old transportation habits. The foresight of the likely changes and getting ready for them smooths the adaptation process. Walker and Marchau [60] creates a framework for future actions that allows for adaptations over time as knowledge about performance and acceptance of the new system accumulates. The paper focuses on the future possibilities that automated transportation will bring in terms of policymaking. It proposes a flexible model that can be shaped under the light of upcoming results of CAVs. Being prepared for future possibilities of new technology is weighty for policymakers. This study could be a handbook for them.

Levin and Boyles [29] state that the widespread use of CAVs is not far away, and they model the benefits it will bring in link to demand, fuel consumption, and traffic. Results of the study show that an increase in transit ridership will reduce personal vehicle trips. New transportation habit predictions are modeled with this research. Having a prediction model for the new transportation habits can determine the regulations and investments in this area.

Liu et al. [33] conducted study based on an extensive literature review and evidence synthesis and is intended as a stimulus for future study and further debate about road infrastructure. The paper focuses on the current situation and possible future changes and proposes a three-phase road infrastructure upgrade plan that evolves. The first phase of the plan spans until 2030 (from the time the study was published which is 2019) where the authors expect level 0 to be dominant. Then phase 2 spans between 2030 and 2050, with the expectation of level 1-2 vehicles to be dominant in the traffic. And finally last phase covers the 2050 s and authors expect to see level 3-4 vehicles dominantly on the road. Roads should be adapted to these changes accordingly.

In a study [55], the authors explore the tensions between democratic experiments and technological ones with a focus on policy for nascent self-driving/automated vehicles. Democratic concerns will arise about how the adaptation process takes. This study considers acceptance as a governance process rather than an individual level. The outcome of this study shows that the more public knows about the CAVs the clearer it will be shaped. Thus, policymakers should consider the importance of education while introducing new technology into daily lives.

Studies show that current road infrastructure is not eligible for fully adapting CAVs into daily traffic. Researchers propose methods to reorganize the road infrastructure. Social life and transportation labor should be considered before launching CAVs on the road. Suggestions for policymakers are brought together in this paper to provide a comprehensive resource.

8. Related work

Our study stands out by synthesizing diverse facets of the challenges hindering the widespread acceptance of CAVs, uniquely combining various perspectives not typically addressed collectively. Unlike many studies that focus on specific subtopics within this domain.

A study [24] explores the gaps in existing research concerning user adoption of CAVs, particularly emphasizing cyber security and privacy concerns. The vulnerability of Federated Learning (FL) to mobile attackers, engaged in model poisoning, is highlighted. The study underscores the absence of validation for robustness against such attacks, proposing exploration of advanced security mechanisms like encryption, localization, behavioral analysis, and clustering to fortify protection against potential CAV-related threats in future research.

In a complementary study, [40] argues for the indispensability of incentives in promoting CAV acceptance. The study, using electric vehicles (EV) and hybrid vehicles as proxies for CAVs due to shared adoption challenges, delves into effective incentive categories. By analyzing incentives implemented in 15 different countries and reviewing literature on various incentive types’ efficacy, the study provides valuable recommendations.

Turning to psychological factors, [45] conducts a literature review on their influence on the acceptability and adoption of CAVs. The authors emphasize the potential impact of individual differences on CAV acceptability, explored further through multiple focus groups. Outcomes from these groups highlight factors such as predictability of CAV behavior, perceived environmental sustainability, a well-defined legal liability framework, and the ability to communicate with other road users. The findings underscore the necessity of tailoring marketing strategies to diverse user characteristics and needs.

Addressing the intersection of technology and acceptance, a study [19] incorporates Internet of Things (IoT)-based technologies in investigating the acceptance of advanced CAV levels (Level 4 and Level 5). The study suggests that IoT-based technology services can expedite CAV adoption by enhancing perceived utility for users. It emphasizes the need to explore business models to identify optimal opportunities in this context.

A comprehensive study by authors [15] delves into user expectations and concerns through focus group sessions across 23 countries. Participants, including both experts in the transport field and novices, voiced diverse concerns, particularly regarding privacy, accident responsibility, and potential increases in maintenance and repair costs. These identified concerns are critical factors that could substantially impede or defer the acceptability of CAVs, demanding careful consideration by stakeholders in the field.

Even though it is not about the adaption of the CAVs, SAE’s [52] classification of levels of automation makes a corner stone for both researchers and policy makers in regulatory aspect. A similar approach is also performed by National Highway Traffic Safety Administration (NHTSA) [37] and BASt levels by German Federal Highway Research Institute [17] where the study [64] makes a comparison of them.

Finally, an independent study that is closest to ours by Bezai et al. [4] where the authors mainly focus on the regulatory and public perception aspect of the adoption. They go over the literature to come up with a suggestion list that will foster the adoption of AVs. Apart from their approach, this paper also includes a classification of common threats that the CAVs face in terms of cyber security. Table 3 shows the comparison of related works and provides an insight of the knowledge gap.

Table 3
Coverage of related work.

Study Cyber security Privacy Road infrastructure Regulation Public perception Incentives

[24] ✓ ✓

[40] ✓

[19,45] ✓

[15] ✓ ✓

[17,37,52] ✓

[4] ✓ ✓ ✓ ✓ ✓

This study ✓ ✓ ✓ ✓ ✓

9. Discussion

CAVs constitute the new era for the transportation industry. They can provide a more sustainable and efficient traveling experience without human interaction. With this technology, human errors in traffic could be eliminated. Moreover, better-planned road trips could be achieved with the help of communication between vehicles and infrastructures. This leads to shorter and more energy-efficient traveling. With all the benefits put aside, certain obstacles prevent CAVs to roll up their wheels on the roads.

The main struggle with this technology is the lack of awareness. This leads people to grow a fear of CAVs which will result in unwillingness to use it. The more people are informed about the better they understand the benefits. This could be achieved either by providing qualified information to them or giving them opportunity to experience the vehicles. According to the survey conducted by Classen et al. [11], acceptance of AVs increases with exposure, regardless of age, as evidenced by their findings. A certain level of training should also be given to vendors who should provide necessary information to the customers. This can help to overcome the dread of CAVs as well as prevent human flaws from a possible cyber attack.

Another issue is cyber attacks that target CAVs. The risk of cybercrime against computational devices is no surprise since the day they entered our lives. CAVs will not be any exception to that. Waiting for a device to become fully secure, or even claiming a device to be fully secure is impractical. Sharma and Kumar Awasthi [51], deals with security and forensics on IoT. They classify literature into three categories: physical level, network level, and cloud level. Their study focuses on extraction techniques and legal issues for forensics and they suggest possible research areas. CAVs can make their way on the road and keep evolving as new attack & defense mechanisms develop. During the categorization of the threats in Table 2, it is important to acknowledge that both the identified threats and the categorization process itself are in their early stages of development. This area presents an open opportunity for future investigation and exploration.

Additionally, privacy protection is an important aspect of CAV development. Attacking vehicles to steal data can be harder to detect and prevent. Data protection should be considered in the early stages of the implementation. Data protection methods like anonymization and encryption should be incorporated in an efficient way to support CAVs needs. It is important for government to regulate privacy.

Regulations cannot be beared on a single country. Due to the nature of data transfer, regulations should be held in a way to protect users all over the world. Data transfer and process are regulated in some regions, same methods can also be applied in the transportation industry. A possible regulation model should be complete in a sense that covers different dependencies mentioned in the Section 6 at once. The CRF model that Khan et al. [26] developed for quantitatively analyzing the problem using SFM might be a possible starting point.

Having autonomously functioning vehicles will have an impact on transportation habits. Drivers might need to look for another career. Roadways and traffic labels might need to be adjusted. Both demographic and physical changes are likely. Preparing for the foreseen changes is necessary to smooth the transformation.

Table 4 summarizes current challanges and recommendations for them. After climbing over the obstacles, a new chapter for transportation is not far away. All these adjustments of technology exist to ease daily life. Adopting CAVs do not only alleviate the burden of driving but also enhance our quality of life and advance our society.

Table 4
Current challenges and recommendations.

Challenge Recommendations References

Lack of awareness Education programs for better understanding the technology behind CAVs can degrade inexperienced road users. It should held mandatory for both CAV users and vendors. This would depreciate number of human errors. [31,32]

Cyber attacks It is an ongoing process for whole ecosystem. Attacks and mitigation techniques related to CAVs are brought together in Table 2. Mandatory disclosure of the vulnerabilities would benefit every stakeholder. Table 2, [10]

Regulation model A comprehensive regulation model that covers international needs should be implemented. A supra-state organisation like EU, can step out to come to a consensus. Different stakeholders should be taken into consideration as well. [2,12,16,26,28]

Social and infrastructural readiness Extensive use of CAVs will create demographic changes. Decision makers and sociologist should study in this issue. Also the infrastructure of roads and components of it should be improved to serve on behalf of CAVs. [33,55,60]

Challenge	Recommendations	References
Lack of awareness	Education programs for better understanding the technology behind CAVs can degrade inexperienced road users. It should held mandatory for both CAV users and vendors. This would depreciate number of human errors.	[31,32]
Cyber attacks	It is an ongoing process for whole ecosystem. Attacks and mitigation techniques related to CAVs are brought together in Table 2. Mandatory disclosure of the vulnerabilities would benefit every stakeholder.	Table 2, [10]
Regulation model	A comprehensive regulation model that covers international needs should be implemented. A supra-state organisation like EU, can step out to come to a consensus. Different stakeholders should be taken into consideration as well.	[2,12,16,26,28]
Social and infrastructural readiness	Extensive use of CAVs will create demographic changes. Decision makers and sociologist should study in this issue. Also the infrastructure of roads and components of it should be improved to serve on behalf of CAVs.	[33,55,60]

10. Conclusion

Current transportation habits are not applicable in today’s world. CAVs provide sufficient improvements in the transportation industry. However, there are certain stumbling blocks in its far-flung use. The obstacles on the widespread use of CAVs should be defined clearly in order to provide a solution. Existing literature is reviewed and provided in this paper to achieve this purpose. Findings are structured under five sub-topics to create an easy-to-read structure: public perception, common attacks, privacy, regulations, and transportation habits.

This paper provides an extensive resource for naming the existing obstacles. Providing solutions for those problems hopefully opens the doors for CAVs. The notable results of this review suggests the following: Public acceptance could be achieved by education programs. Mitigation techniques for cyber attacks can improve the security perception. Data protection methods should be used to provide privacy for the users. Regulation on privacy protection and road infrastructure necessitates confidentiality. New transportation habits must be encouraged.

A researcher could chose any of the aforementioned sections as a starting point depending on their profession. The broadly processed sub-topics could be elaborated. Ultimately, the widespread adoption of CAVs are expected to transform the way people travel, making transportation more efficient, safer, and more convenient for everybody.

Footnotes

Conflict of interest

The authors have no conflict of interest to report.

References

Atmaca

U.I.

Maple

Dianati

, Emerging privacy challenges and approaches in CAV systems, in: Living in the Internet of Things (IoT 2019), 2019, pp. 1–9. doi:https://doi.org/10.1049/cp.2019.0141.

Baker

Theissen

C.M.

Vakil

, Connected and autonomous vehicles: A cross-jurisdictional comparison of regulatory developments, The Journal of Robotics, Artificial Intelligence & Law 3(4) (2020), 249–273, https://heinonline.org/HOL/LandingPage?handle=hein.journals/rail3&div=38&id=&page=.

Bansal

Kockelman

K.M.

Singh

, Assessing public opinions of and interest in new vehicle technologies: An Austin perspective, Transportation Research Part C: Emerging Technologies 67 (2016), 1–14. doi:https://doi.org/10.1016/j.trc.2016.01.019.

Bezai

N.E.

Medjdoub

Al-Habaibeh

Chalal

M.L.

Fadli

, Future cities and autonomous vehicles: Analysis of the barriers to full adoption, Energy and Built Environment 2(1) (2021), 65–81. doi:https://doi.org/10.1016/j.enbenv.2020.05.002.

Braun

Clarke

, Using thematic analysis in psychology, Qualitative Research in Psychology 3(2) (2006), 77–101. doi:https://doi.org/10.1191/1478088706qp063oa.

Buchholz

, Charted: Autonomous driving is racing ahead. https://www.weforum.org/agenda/2023/02/charted-autonomous-driving-accelerating-mobility/.

Cao

Xiao

Cyr

Zhou

Park

Rampazzi

Chen

Q.A.

Mao

Z.M.

, Adversarial sensor attack on LiDAR-based perception in autonomous driving, in: Proceedings of the 2019 ACM SIGSAC Conference on Computer and Communications Security, CCS’19, Vol. 63, Association for Computing Machinery, New York, NY, USA, 2019, pp. 2267–2281. doi:https://doi.org/10.1145/3319535.3339815.

Chellapandi

V.P.

Yuan

Zak

S.H.

Wang

, A survey of federated learning for connected and automated vehicles, 2023. https://arxiv.org/abs/2303.10677.

Chikaraishi

Khan

Yasuda

Fujiwara

, Risk perception and social acceptability of autonomous vehicles: A case study in Hiroshima, Japan, Transport Policy 98 (2020), 105–115. doi:https://doi.org/10.1016/j.tranpol.2020.05.014.

10.

Choi

Fershtman

Gandal

, Network security: Vulnerabilities and disclosure policy, The Journal of Industrial Economics 58 (2010), 868–894. doi:https://doi.org/10.1111/j.1467-6451.2010.00435.x.

11.

Classen

Sisiopikub

V.P.

Masona

J.R.

Yangb

Hwangboa

S.-W.

McKinneyb

, Experience of drivers of all age groups in accepting autonomous vehicle technology, Journal of Intelligent Transportation Systems (2023), 1–17. doi:https://doi.org/10.1080/15472450.2023.2197115.

12.

D’Amato

Dancel

Pilutti

Tellis

Frascaroli

Gerdes

J.C.

, Exceptional driving principles for autonomous vehicles, Journal of Law and Mobility 2022 (2022), https://repository.law.umich.edu/jlm/vol2022/iss1/2.

13.

Davis

F.D.

, Perceived usefulness, perceived ease of use, and user acceptance of information technology, MIS Quarterly 13(3) (1989), 319–340. doi:https://doi.org/10.2307/249008.

14.

Dow

, What are the six SAE levels of self-driving cars? 2023. https://www.topgear.com/car%20news/what-are-sae-levels-autonomous-driving-uk.

15.

Duboz

Mourtzouchou

Grosso

Kolarova

Cordera

Nägele

Alonso Raposo

Krause

Garus

Eisenmann

dell’Olio

Alonso

Ciuffo

, Exploring the acceptance of connected and automated vehicles: Focus group discussions with experts and non-experts in transport, Transportation Research Part F: Traffic Psychology and Behaviour 89 (2022), 200–221. doi:https://doi.org/10.1016/j.trf.2022.06.013.

16.

Fagnant

D.J.

Kockelman

, Preparing a nation for autonomous vehicles: Opportunities, barriers and policy recommendations, Transportation Research Part A: Policy and Practice 77 (2015), 167–181. doi:10.1016/j.tra.2015.04.003.

17.

Gasser

T.M.

Westhoff

, BASt-study: Definitions of automation and legal issues in Germany, Automation Workshop, 2012, Online. https://onlinepubs.trb.org/onlinepubs/conferences/2012/Automation/presentations/Gasser.pdf.

18.

Girka

Terziyan

Gavriushenko

Gontarenko

, Anonymization as homeomorphic data space transformation for privacy-preserving deep learning, Procedia Computer Science 180 (2021), 867–876. doi:https://doi.org/10.1016/j.procs.2021.01.337.

19.

Harrison

Shepherd

S.P.

Chen

, Modelling uptake sensitivities of connected and automated vehicle technologies, International Journal of System Dynamics Applications 10(2) (2021), 88–106. doi:https://doi.org/10.4018/ijsda.2021040106.

20.

Söffker

, Identification of human driver critical behaviors and related reliability evaluation in real time, in: Proceedings of the 32nd European Safety and Reliability Conference (ESREL 2022), Research Publishing, Singapore, 2022, pp. 2112–2119.

21.

Meng

, Survey on cyber security of CAV, in: 2017 Forum on Cooperative Positioning and Service (CPGPS), 2017, pp. 351–354. doi:https://doi.org/10.1109/CPGPS.2017.8075153.

22.

Hilgarter

Granig

, Public perception of autonomous vehicles: A qualitative study based on interviews after riding an autonomous shuttle, Transportation Research Part F: Traffic Psychology and Behaviour 72 (2020), 226–243. doi:https://doi.org/10.1016/j.trf.2020.05.012.

23.

Hillebrand

, The role of fear and trust when disclosing personal data to promote public health in a pandemic crisis, in: Innovation Through Information Systems, Ahlemann

Schütte

Stieglitz

, eds, Springer International Publishing, Cham, 2021, pp. 247–262. ISBN 978-3-030-86790-4. doi:https://doi.org/10.1007/978-3-030-86790-4_18.

24.

Hussain

Rani

Chouhan

Gaur

U.S.

, Cyber security and privacy of connected and automated vehicles (CAVs)-based federated learning: Challenges, opportunities, and open issues, in: EAI/Springer Innovations in Communication and Computing Book Series, Springer, 2022. doi:https://doi.org/10.1007/978-3-030-85559-8_11.

25.

H.J.

Choi

S.Y.

Woo

Lee

D.H.

, Vulnerabilities of Android OS-based telematics system, Wireless Personal Communications 92 (2017), 1512–1530. doi:https://doi.org/10.1007/s11277-016-3618-9.

26.

Khalid Khan

Shiwakoti

Stasinopoulos

Warren

, Modelling cybersecurity regulations for automated vehicles, Accident Analysis & Prevention 186 (2023), 107054. doi:https://doi.org/10.1016/j.aap.2023.107054.

27.

Kyriakidis

Happee

de Winter

J.C.F.

, Public opinion on automated driving: Results of an international questionnaire among 5000 respondents, Transportation Research Part F: Traffic Psychology and Behaviour 32 (2015), 127–140. doi:https://doi.org/10.1016/j.trf.2015.04.014.

28.

Lee

Hess

, Regulations for on-road testing of connected and automated vehicles: Assessing the potential for global safety harmonization, Transportation Research Part A: Policy and Practice 136 (2020), 85–98. doi:https://doi.org/10.1016/j.tra.2020.03.026.

29.

Levin

M.W.

Boyles

S.D.

, Effects of autonomous vehicle ownership on trip, mode, and route choice, Transportation Research Record: Journal of the Transportation Research Board 2493 (2019), 29–38. doi:https://doi.org/10.3141/2493-04.

30.

Liu

Shi

, 4C: A computation, communication, and control co-design framework for CAVs, IEEE Wireless Communications 28(4) (2021), 42–48. doi:https://doi.org/10.1109/MWC.201.2000512.

31.

Liu

, Exploring public acceptance of connected and autonomous vehicles with a focus on cyber security and privacy risks, PhD thesis, University of Huddersfield, Huddersfield, UK, 2021.

32.

Liu

Nikitas

Parkinson

, Exploring expert perceptions about the cyber security and privacy of connected and autonomous vehicles: A thematic analysis approach, Transportation Research Part F: Traffic Psychology and Behaviour 75 (2020), 66–86. doi:https://doi.org/10.1016/j.trf.2020.09.019.

33.

Liu

Tight

Sun

Kang

, A systematic review: Road infrastructure requirement for Connected and Autonomous Vehicles (CAVs), Journal of Physics: Conference Series 1187(4) (2019), 042073. doi:https://doi.org/10.1088/1742-6596/1187/4/042073.

34.

Maeng

Kim

Cho

, Consumers’ attitudes toward information security threats against connected and autonomous vehicles, Telematics and Informatics 63 (2021), 101646. doi:https://doi.org/10.1016/j.tele.2021.101646.

35.

Mercedes-Benz Group, Mercedes-Benz world’s first automotive company to certify SAE level 3 system for U.S. market, 2023. https://web.archive.org/web/20231210210705/https://group.mercedes-benz.com/innovation/product-innovation/autonomous-driving/drive-pilot-nevada.html.

36.

Naessens

Sandikkaya

M.T.

Lapon

Verslype

Verhaeghe

Nigusse

De Decker

, Privacy policies, tools and mechanisms of the future, in: iNetSec 2009 – Open Research Problems in Network Security, Springer, 2009, pp. 125–138. doi:https://doi.org/10.1007/978-3-642-05437-2_12.

37.

National Highway Traffic Safety Administration, Levels of automation. https://www.nhtsa.gov/document/levels-automation.

38.

Nordhoff

de Winter

Payre

van Arem

Happee

, What impressions do users have after a ride in an automated shuttle? An interview study, Transportation Research Part F: Traffic Psychology and Behaviour 63 (2019), 252–269. doi:https://doi.org/10.1016/j.trf.2019.04.009.

39.

Parkinson

Ward

Wilson

Miller

, Cyber threats facing autonomous and connected vehicles: Future challenges, IEEE Transactions on Intelligent Transportation Systems 18(11) (2017), 2898–2915. doi:https://doi.org/10.1109/TITS.2017.2665968.

40.

Penmetsa

Dhondia

Adanu

E.K.

Harper

Nambisan

Jones

, Incentives to encourage the adoption of connected and automated vehicles: Lessons learned from hybrid-electric vehicle incentive programs, Future Transportation 3(3) (2023), 986–995. doi:https://doi.org/10.3390/futuretransp3030054.

41.

Petit

Shladover

S.E.

, Potential cyberattacks on automated vehicles, IEEE Transactions on Intelligent Transportation Systems 16(2) (2015), 546–556. doi:https://doi.org/10.1109/TITS.2014.2342271.

42.

Petit

Stottelaar

Feiri

Kargl

, Remote attacks on automated vehicles sensors: Experiments on camera and LiDAR, in: Proceedings of Black Hat Europe 2015. https://www.blackhat.com/docs/eu-15/materials/eu-15-Petit-Self-Driving-And-Connected-Cars-Fooling-Sensors-And-Tracking-Drivers-wp1.pdf.

43.

Philipsen

S.G.

Andersen

Singh

, Threats and attacks to modern vehicles, in: 2021 IEEE International Conference on Internet of Things and Intelligence Systems (IoTaIS), 2021, pp. 22–27. doi:https://doi.org/10.1109/IoTaIS53735.2021.9628576.

44.

Piao

McDonald

Hounsell

Graindorge

Malhene

, Public views towards implementation of automated vehicles in urban areas, Transportation Research Procedia 14 (2016), 2168–2177. doi:https://doi.org/10.1016/j.trpro.2016.05.232.

45.

Post

J.M.M.

Veldstra

J.L.

Ünal

A.B.

, Acceptability and acceptance of connected automated vehicles: A literature review and focus groups, in: Proceedings of the 5th International Conference on Computer-Human Interaction Research and Applications (CHIRA 2021), 2021, pp. 223–231. doi:https://doi.org/10.5220/0010719200003060.

46.

Psiaki

M.L.

Humphreys

T.E.

, GNSS spoofing and detection, Proceedings of the IEEE 104(6) (2016), 1258–1270. doi:https://doi.org/10.1109/JPROC.2016.2526658.

47.

Pyrialakou

Gkartzonikas

Gatlin

J.D.

Gkritza

, Perceptions of safety on a shared road: Driving, cycling, or walking near an autonomous vehicle, Journal of Safety Research 72 (2020), 249–258. doi:https://doi.org/10.1016/j.jsr.2019.12.017.

48.

Salonen

A.O.

Haavisto

, Towards autonomous transportation. Passengers’ experiences, perceptions and feelings in a driverless shuttle bus in Finland, Sustainability 11(3) (2019), 588. doi:https://doi.org/10.3390/su11030588.

49.

Schoettle

Sivak

, Public opinion about self-driving vehicles in China, India, Japan, the US, the UK, and Australia, Technical Report, 103139, University of Michigan, Transportation Research Institute, Ann Arbor, Michigan, 2014.

50.

Sharma

Mishra

, Ranking preferences towards adopting autonomous vehicles based on peer inputs and advertisements, Transportation 50(6) (2023), 2139–2192. doi:https://doi.org/10.1007/s11116-022-10304-w.

51.

Sharma

Kumar Awasthi

, IoT forensics in ambient intelligence environments: Legal issues, research challenges and future directions, Journal of Ambient Intelligence and Smart Environments Pre-press (2023), 1–38. doi:https://doi.org/10.3233/AIS-220511.

52.

Shuttleworth

, SAE standards news: J3016 automated-driving graphic update, Technical Report, SAE International, 2019.

53.

Sitawarin

Bhagoji

A.N.

Mosenia

Mittal

Chiang

, Rogue Signs: Deceiving Traffic Sign Recognition with Malicious Ads and Logos, 2018, https://arxiv.org/abs/1801.02780 .

54.

Stempel

, Tesla, Musk sued by shareholders over self-driving safety claims, 2023. https://web.archive.org/web/20231210211039/https://www.reuters.com/legal/tesla-musk-sued-by-shareholders-over-self-driving-safety-claims-2023-02-27/.

55.

Stilgoe

Cohen

, Rejecting acceptance: Learning from public dialogue on self-driving vehicles, Science and Public Policy 48 (2021), 849–859. doi:https://doi.org/10.1093/scipol/scab060.

56.

Sun

Zhang

, A survey on cyber-security of connected and autonomous vehicles (CAVs), IEEE Transactions on Intelligent Transportation Systems 23(7) (2022), 6240–6259. doi:https://doi.org/10.1109/tits.2021.3085297.

57.

The MITRE Corporation, ATT&CK. https://attack.mitre.org/.

58.

Treat

J.R.

, A study of precrash factors involved in traffic accidents, HSRI Research Review 10(6) (1980), 35.

59.

Venkatesh

Morris

M.G.

Davis

G.B.

Davis

F.D.

, User acceptance of information technology: Toward a unified view, MIS Quarterly 27(3) (2003), 425–478. doi:https://doi.org/10.2307/30036540.

60.

Walker

W.E.

Marchau

V.A.

, Dynamic adaptive policymaking for the sustainable city: The case of automated taxis, International Journal of Transportation Science and Technology 6 (2017), 1–12. doi:https://doi.org/10.1016/j.ijtst.2017.03.004.

61.

Liao

Wang

J.-W.

, Analysis of consumer attitudes towards autonomous, connected, and electric vehicles: A survey in China, Research in Transportation Economics 80 (2020), 100828. doi:https://doi.org/10.1016/j.retrec.2020.100828.

62.

Xiong

Zhao

Guo

Yang

, Edge-assisted privacy-preserving raw data sharing framework for connected autonomous vehicles, IEEE Wireless Communications 27(3) (2020), 24–30. doi:https://doi.org/10.1109/mwc.001.1900463.

63.

Yang

Feng

Zhang

Shi

, A novel data fusion algorithm to combat false data injection attacks in networked radar systems, IEEE Transactions on Signal and Information Processing over Networks 4(1) (2018), 125–136. doi:https://doi.org/10.1109/TSIPN.2018.2790361.

64.

Yang

Jiang

Zhao

Cao

Xie

Zhongyang

Jiao

Wang

Zhang

, Intelligent and connected vehicles: Current status and future perspectives, Science China Technological Sciences 61 (2018). doi:https://doi.org/10.1007/s11431-017-9338-1.

65.

Zhang

Shi

, Controlling a car through OBD injection, in: 2016 IEEE 3rd International Conference on Cyber Security and Cloud Computing (CSCloud), 2016, pp. 26–29. doi:https://doi.org/10.1109/CSCloud.2016.42.

66.

Zhao

Chen

, Minimum delay routing for connected and autonomous vehicles communications, in: 2020 12th International Symposium on Communication Systems, Networks and Digital Signal Processing (CSNDSP), 2020, pp. 1–5. doi:https://doi.org/10.1109/CSNDSP49049.2020.9249520.

Study	Cyber security	Privacy	Road infrastructure	Regulation	Public perception	Incentives
[24]	✓	✓
[40]						✓
[19,45]					✓
[15]		✓			✓
[17,37,52]				✓
[4]	✓	✓	✓	✓	✓
This study	✓	✓	✓	✓	✓

A survey on obstacles to the widespread use of connected and automated vehicles

Abstract

Keywords

1. Introduction

2.1. Search strategy

2.2. Writing process

3. Public perception

3.1. Perception model

3.2. Perception surveys

3.3. Suggestions

4. Common attacks and mitigation techniques

4.1. Existing attack attempts

4.3. Technical improvements

5. Privacy

6. Regulations

7. Transportation habits

8. Related work

Table 3 Coverage of related work. Study Cyber security Privacy Road infrastructure Regulation Public perception Incentives [24] ✓ ✓ [40] ✓ [19,45] ✓ [15] ✓ ✓ [17,37,52] ✓ [4] ✓ ✓ ✓ ✓ ✓ This study ✓ ✓ ✓ ✓ ✓

Footnotes

Conflict of interest

References

Table 3
Coverage of related work.

Study Cyber security Privacy Road infrastructure Regulation Public perception Incentives

[24] ✓ ✓

[40] ✓

[19,45] ✓

[15] ✓ ✓

[17,37,52] ✓

[4] ✓ ✓ ✓ ✓ ✓

This study ✓ ✓ ✓ ✓ ✓