Abstract
Autonomous vehicles have increasingly become popular in many countries because of the advantages that come with them. They could particularly be used by those that do not have the capabilities to drive or urban residents amongst others. Specifically, human errors have necessitated the development of intelligent self-driving cars that could help in avoiding accidents. Through computer technology, these vehicles are now capable of moving from one point to another without the intervention of human beings. However, there are also concerns about the privacy of their use, cybersecurity and safety. This literature review highlights the increased adoption of autonomous vehicles and the concerns associated with them. It also offers information on the attempts made to address the privacy, cybersecurity and safety issues in many countries.
?
Introduction
Autonomous vehicles, also referred to as self-driving cars are vehicles that can sense their environments and move safely without the intervention or input of human beings. They use a lot of sensors to learn their surroundings such as the navigation paths, signage and obstacles that should be considered in movement (Seshan and Maitra, 2014). Accordingly, autonomous denotes the fact that it is self-governing. Moreover, the autonomous control means that the gadget in this case the vehicle can satisfactorily perform under critical uncertainties in the environment. These gadgets also have the ability to compensate for the system failures in such an environment without the intervention of an external entity. Wood et al (2012) reaffirm the use of the term autonomous and argue that it is more widespread and familiar to the users or the public. Apart from the issues concerning privacy, cybersecurity and safety, it has been established that there are legal issues that affect automated driving. The legal and ethical concerns that affect the technological feasibility of autonomous vehicles have been discussed with differing opinions (Martínez-Díaz and Soriguera, 2018). It is against this backdrop that this literature review seeks to evaluate the adoption of autonomous vehicles.
Figure 1: Autonomous Vehicle Technologies(Canis, 2018)
Privacy
Autonomous vehicles rely on sensors, maps and other instruments from which they obtain information and use them to ensure that the vehicle can be operated safely (West, 2016). However, there are reservations on whoever exercises control over this information and the manner it is used (Boeglin, 2015). This is because there is a number of issues that concern the privacy of such information. In addition, there are concerns about the specific reasons for the collection of such information, type, accessibility and permitted timeline for which it should be stored (Glancy, 2012). It is argued that such communication allows the information to be passed over to another but at the same time exposing exact place and movement of a car to external or unrecognized networks. As a result of this, Schoonmaker (2016) has argued that securing the data about the location of a car may not be enough. The reason for this is that the permission of the customer may not have been obtained at the time the autonomous vehicle is bought and used. Further, Dhar (2016) has argued that the use of event data recorders (EDRs) in autonomous vehicles to confirm the causes of events such as accidents may compromise the privacy of the users. Under this, it is argued that such information may be illegally transferred to third parties such as insurers and used against drivers.
Again, Glancy (2012) states that there is risk to informational privacy of a person as such data may be used to harass people. This can be done through promotions and marketing that is designed to unlawfully obtain the identities of people with the aim of predicting their activities. It could also be used to bring together information and manipulate others (Glancy, 2012). Even more significant is the fact that anonymous information about users can be made known by revealing the exact names. Indeed, Gambs, Killijian and del Prado Cortez (2014) confirm that data obtained about a location can be used to locate the movement of vehicles and the actual users. Therefore, Glancy (2012) states that the wireless network which is connected to the autonomous vehicles can assist different agencies in surveying a person. The result is that this goes against the principle of autonomy and may subject individuals to mental manipulation and harassment.
Different countries have made laws that protect their privacy and how the data obtained from autonomous vehicles may be collected, stored or stored. Crane, Logue and Pilz (2016) have argued that the SPY Car Act allows the US National Highway Traffic Safety Administration ((NHTSA) to protect the access and use of driving data in all cars that are made for sale in the United States. Under this law, all vehicles must offer the owners or those who hire them to stop collection of information apart from that which is required to ensure safety or investigations of accidents. Further, the makers are prohibited from using this information for marketing and promotions without the express permission of owners. Similarly, the European Union (EU) has put in place steps to manage the privacy of data within the region. In support of this, there is already a measure under which permission may be obtained from users. It is also added that users have the right to seek to be forgotten to the system or ask for subsequent information regarding the decisions that companies make about the use of their data.
Cybersecurity
Already, conventional vehicles suffer from cybersecurity concerns as long as there are automated tools within them (Hancock, Nourbakhsh and Stewart, 2019). This is because there is chance that unauthorized or unintended attacks on its computer operating systems may impact on the way they are run. König and Neumayr (2017) have evaluated the perception of users and expounded on the reasons why they tend to resist the self-driving vehicles. While they have increasingly become acceptable, they are also confusing to the users. This is because of issues concerning the probability that these technologies may be abused in terms of compromising the security of the data. Most studies indicate that the concerns of owners and users concerning cybersecurity stem from the possibilities of hacking of software and misuse of such vehicles (Kyriakidis, Happee and de Winter, 2015). Under this view, the fear is that hackers now have the ability to control or manage vehicles by taking advantage of the wireless networks that improve the connectivity of the car with the external environment. Specifically, autonomous vehicles have the ability to store and pass over data and transactions. Therefore, Lee (2017) states that such information can be used by hackers for illegal or financial gains. Additionally, they could be used by criminal elements such as extremists and terrorists to direct attacks or cause physical harm to users. In any case, Lee (2017) confirms this position by stating that autonomous vehicles become more vulnerable because the computers have better control over the movement of such a vehicle and the driver may not act if an incident were to occur.
Further, the input of fake messages and spoofing of the international navigation satellite systems have increasingly become an emerging threat to the use of autonomous vehicles. Bagloee et al (2016) point out that the manipulation of the data within these systems can affect the critical gadgets that ensure the safety of these cars or lead to accidents. Therefore, this author affirms that there may need to continuously update the software that cars use. Bagloee et al (2016) also propose the changing of the present standardized security measures that have been in use in protecting the vehicles from cyber-attacks. Schellenkens (2016) evaluated the challenges posed by the hacking of these cars and the need to abide by the set regulations against cyber attacks. This is against the backdrop of successful hacking of a vehicle in the United States after which the hackers took over the crucial functions including the braking and engine operations. This raises the question on the extent to which the manufacturers of autonomous vehicles go ensure that they counteract threats posed by cybersecurity, especially the need to enact laws to regulate the same.
Most countries have come up with guidelines on cybersecurity that are not necessarily mandatory but contain the best practices that can guarantee the security of autonomous vehicles. For instance, the NHTSA (2018) provides ways of assessing and monitoring the possible cyber threats that may occur within the transport systems. This is an affirmation of the efforts made to raise and improve awareness on the cybersecurity risks that the manufacturers and the collaborating software developers are likely to meet during the manufacturing of the self-driving cars. Moreover, the American SPY Car Act also enhances the controls placed on cybersecurity in all cars that call for assessing the vehicle’s software using the industry best practices. Accordingly, Taeihagh and Lim (2019) have pointed out that different nations have ways of handling cybersecurity issues that affect autonomous vehicles. For example, Japan has demonstrated a movement towards enacting a law that will govern the use of these vehicles. However, this has not been actualized yet as the country still does not have concrete strategies for accomplishing such a task.
Figure 2: Entry Points for Vehicle Hacking (Canis, 2018)
Safety
Most motor vehicle accidents in the world can be attributed to human errors (Smith, 2013). Therefore, it may at times be argued that using autonomous vehicles has the probability of reducing or eliminating the significant cause of motor vehicle accidents. While these vehicles could also enhance the human drivers concerning perceptions, decision-making and implementation, it may raise safety issues. Collingwood (2017) states that the occupants of a car may avoid using safety protections like seat belts. This may increase the chances of accidents occurring. Moreover, the mere assumption that human error has been eliminated does not overrule the chances of malfunctioning of the machine or its potential errors. It implies that the technological growth of autonomous vehicles also increases the probability of technical errors that can affect the safety of users. To demonstrate this, Banks, Plant and Stanton (2018) offers the example of the deadly crash of the Tesla autopilot that occurred in 2016 that confirms that the technical ability of the autonomous vehicles may be impaired.
Coca-Vila (2018) has also raised safety concerns about the use of autonomous vehicles regarding the way in which they should be programmed with algorithms that can adequately react whenever there are unavoidable incidents. Here, it is stated that the absence of human drivers to man the machines makes it difficult to assign liability if an accident occurs. Additionally, this leads to a situation where the regulations about these vehicles may be subjected to moral judgments. Accordingly, Coca-Vila (2018) has noted that using economic considerations as the cornerstone of developing self-driving vehicles and emphasis on the safety of occupants goes against the value of safety laws. To confirm the gravity of this matter, it is also noted that Bansal, Kockelman and Singh (2016) point out that there is little acceptance of autonomous vehicles by the public because of the safety concerns about this technology. Part of the reason for this is that there is little agreement on the real level of safety, or how this can be determined in the case of self-driving cars.
Different countries have their own national safety standards that regulate transportation and the behaviour of the drivers. Accordingly, Stone (2018) clearly notes that the US Self Drive Act sets out the duties of the NHTSA as the regulatory body that governs the activities of the autonomous vehicles. This includes matters concerning safety strategies and designs of the cars that must only be identical to the ones prescribed by the laws. Increasingly, the role of federal governments has increased over and above that of the state governments. While these restrictions on safety have become common amongst governments, it is noteworthy that they are also less stringent such that it allows for room for innovation and development of safer autonomous vehicles.
Koopman and Wagner (2017) have evaluated the safety of autonomous vehicles and advocates for an interdisciplinary approach that can adequately guarantee this important aspect of technology. Under such an intelligent transport system, there is need to have a multi-disciplinary that cuts across all its functions. This includes creating vehicles that can tolerate faults and are resilient. Moreover, it is proposed that human beings that use conventional vehicles should cooperate on the road transport systems. Further, Koopman and Wagner (2017 also recommend that the systems that are used in the driving experience should be structured and according to the set regulations. Under this, it then argued that the designs that are created and deployed must incorporate the safety concerns during the designing and manufacturing phases of these cars. On the other hand, Fernandes and Nunes (2012) propose the use of platooning information management practices that can enable safe and consistent operations of autonomous vehicles. Part of this is to use new algorithms that can support faster communication and stability between the vehicle and the network under which it runs. Pursuant to this, the emergency responses may be responded to in a timely manner without delays. It could also be used to update data that can be used to enhance safety through prompt responses to both foreseen and unforeseen eventualities.
Countries Safety Privacy Cybersecurity
US Light control Control Control
UK Light control Light control Adaptation
Australia Light control, Adaptation Light control, Adaptation No response
EU Light control Control Light control
Germany Control Light control, Control No response
China Prevention, Light control Control Control
Singapore Control Control Control, Adaptation
Japan Prevention Control No response
South Korea No response Light control No response
Table 1: Summary of Selected Government Strategies Adopted (Taeihagh and Lim, 2019).
Conceptual Framework
The autonomous vehicles have many advantages but these come with enormous risks that are associated with privacy, cybersecurity and safety of the cars. First, the provision, use and storage of data have presented challenges as there may be unauthorized access to it by illegal elements or criminals, especially when the geographic location of the vehicle can be traced (Neumann, 2016). Further, Taeihagh and Lim (2019) have identified cybersecurity issues such as hacking as some of the challenges that adoption of the autonomous vehicles portent. Despite this, there are limited regulatory practices and laws that govern these including the safety of people. This literature review discusses the use of autonomous vehicles under three subtopics namely privacy, cybersecurity and safety. However, it is argued that their technological feasibility should adequately address the anticipated user-vehicle interactions concerning the need to uphold privacy and safety as well as cybersecurity. Therefore, the success of adoption of autonomous vehicle depends on effectiveness in their use in terms of privacy, cybersecurity and safety. For example, lack of privacy, loss of critical data and accidents due to system failure is likely to scare away purchasers of self-driving cars.
Research Gap
Multiple research articles have been published in the academic literature that shows the technological advancements of the autonomous vehicles. However, there is limited information on the induced disruptions and how policies can be applied to support or address the disruptive impacts are fairly limited (Bagloee et al, 2016). Even more interesting is the fact that there is no study in academic literature that critically analyzes the state of autonomous vehicles from a joint perspective that focuses on the possible policy interventions on the disruptive effects. It is for this reason that this paper seeks to determine the current position, the impacts of the uptake of the self-driving cars and what can be done to address the identified issues. Additionally, it is noted that the increased levels of vehicle automation pose challenges concerning its use. Particularly, the pre-deployment policies revolve around the privacy, cybersecurity and safety. The tracing technology in these vehicles may have significant impact on the privacy of people and manufacturers can be held liable for the breach (Collingwood, 2017). Similarly, the absence of a cybersecurity system may hinder the storage of onboard data or its sharing. All these factors have raised questions amongst researchers on who should regulate such factors; whether it is the federal or state governments or international legal bodies on the one hand and the manufacturers on the other.
Conclusion
It is evident that the desire for self-driving cars is already here and improvement in technology will make them even better in future. However, the capability of the autonomous vehicle is subjected to issues associated with security, privacy and safety. These vehicles will be controlled by systems which have been proven to have limitation in the past and thus it is important to consider production processes that will address the problem associated with technology issues. It is also important for the governments and international agencies to consider making laws that would strengthen the safety, data privacy and security of the autonomous vehicle.
