Research on the Impact of Autonomous Vehicles on the Environment and Sustainability of Transport Companies

Abstract

The significance of sustainable development and the increasing acknowledgment of climate change-related issues necessitate modifications in the transportation sector, the most polluting of all economic sectors, which is highly reliant on fossil fuels. The transportation sector significantly adversely affects the environment and human health. However, there is still a lack of research in scientific literature confirming the effectiveness of autonomous vehicles. The aim of the article is to examine the experience of autonomous vehicles in society. The primary attitudes of individuals regarding autonomous vehicles were assessed. The research technique involves a statistical examination of elements affecting client apprehensions over the usage of autonomous vehicles (AVs). Results of the research show the impact of autonomous vehicles on the quality of service of transport companies.

1. Introduction

The importance of sustainable development and the growing recognition of the problems caused by climate change requires changes in the field of transportation, the most polluting economic sector, which heavily depends on fossil fuels. The transportation sector has a considerable negative impact on the environment and human health.

The article’s goal is to examine the societal effects—both negative and positive—of autonomous automobiles. The extensive use of self-driving cars would lead to a decrease in the number of cars on the roads, hence reducing CO₂ emissions and road fatalities.

The predominant perspectives of individuals concerning autonomous vehicles were evaluated. The research method entails a statistical analysis of factors influencing client concerns over the use of autonomous vehicles (AVs). Each ton of CO₂ emitted exacerbates global warming. Therefore, any reduction in emissions can mitigate its progression. However, transportation contributes around 25% of the EU’s total emissions of carbon dioxide (CO₂), the main greenhouse gas, and depletes one of the most important non-renewable natural resources [1]. Recognizing the substantial and irreversible detrimental effects of CO₂ on human health and global ecosystems, scientists and business professionals are examining this issue to develop technical improvements and explore new solutions. EU nations approved the objective of reducing greenhouse gas emissions by approximately 55% by the year 2030 and the Vision Zero project, which aspires to eliminate traffic fatalities by 2050. In addition, with the objective of becoming the world’s inaugural climate-neutral continent, as mentioned by Igini [2], the European Commission unveiled the European Green Deal in December 2019. This initiative represents the most ambitious set of policies aimed at providing European residents and businesses with the possibility to benefit from the sustainable shift to a green economy and meet the targets of the Paris Agreement by 2050 [3].

According to this viewpoint, autonomous vehicles (AVs) hold great promise for enabling safer, lighter, and more economical driving while simultaneously encouraging shared and on-demand transportation. Numerous firms are diligently advancing and evaluating autonomous vehicle technology, aspiring for its extensive implementation in the future. The future of autonomous vehicles (AVs) attracts significant attention in both popular opinion and scholarly research due to its potential. Autonomous cars, also called self-driving cars, are equipped with technology that enables them to navigate and operate independently of human intervention using a mix of hardware and software components. The leading concern is the influence of autonomous vehicles on the environment, which is gathering increasing attention. For instance, some researchers [4,5] examined issues connected with the consumption of energy and emissions. The findings presented by Bandeira et al. [6] aim to investigate the environmental effects of connected and automated vehicles (CAVs) by adding various road types with varying features, such as traffic volumes and speed limitations. For the case study, a network made up of three distinct road types—urban, rural, and freeway segments—was taken into consideration. These case studies were selected primarily because they differed in terms of traffic quantities, speed limits, and singularities. The following conclusions could be made thanks to the findings: Taking into account the driving behavior reference values established in the literature, the impact of CAVs on emissions varied greatly depending on the type of road. With up to 10% reductions in emissions, CAVs have been demonstrated to be especially environmentally friendly on national roads. On the other hand, negative effects were demonstrated in the urban corridor. The effects are likewise insignificant at the motorway level, where the V/C ratio is minimal. However, reducing the speed to 90 km/h enables reductions of up to 32% of NO_x and 18% of CO₂. The presence of CAVs had good environmental effects in areas outside of metropolitan areas, and these effects showed a significant linear association with increased market penetration rate (MPR). The presence of CAVs had good environmental effects in areas outside of metropolitan areas, and these effects showed a significant linear association with increased market penetration rate (MPR). MPR 30% and MPR 50% showed the most detrimental effects in the metropolitan areas. The integration of connected autonomous vehicles (AVs) in mixed traffic conditions may result in a 4% increase in CO₂ emissions, but it may also yield an 18% reduction, depending on the type of road, driving conditions, and penetration rate. For these MPRs, total CO₂ and NO_x emissions increased up to 4% and 8% in the urban avenue, respectively.

Greenwald and Kornhauser [7] identified various issues that autonomous vehicles (AVs) could address, including 600 billion driving hours annually with little seat occupancy, 1.3 million yearly worldwide accident fatalities, cars that are parked 96% of the time, and stress associated with driving. Additionally, they noted potential societal opportunities linked to commercial prospects and synergies with public transport services.

Despite the significant potential advantages of autonomous vehicles, their implementation may also produce adverse impacts, necessitating careful management to achieve optimal outcomes [8]. Additionally, assessments conducted by various authors about the accessibility of new technology exhibit considerable variability; however, they often indicate a timeline extending to 2030, assuming that an additional 10 to 20 years will be required to realize a scenario based on autonomous vehicles [9], transportation services, etc.

The technology necessary to attain a significant level of autonomy, which will positively impact the environment, remains under development [10], as autonomous vehicle technology faces limitations in extreme weather conditions, complex urban environments with pedestrians and cyclists, and scenarios involving unpredictable human behavior. It is essential to highlight that the development of new autonomous vehicle technology is costly and may restrict consumer accessibility. The study revealed how important it is for companies to see the speed of autonomous vehicles increase. The study findings suggest that self-driving cars can highlight social disparities.

The primary issue with deploying autonomous vehicles (AVs) is the uncertainty around their safety, the decision-making processes in complicated scenarios, their responses to unforeseen circumstances, and access to consumers.

Severino et al.’s study [11] aimed to increase interest in autonomous vehicles by outlining some of the advantages that come with integrating them into road networks. The article began by outlining the fundamentals of AVs, including their history and development over time, some of their primary applications (such as the transportation of biohazardous materials during pandemics), potential government regulation strategies, and some of the risks associated with AVs. Along with discussing some of the models used for simulations, the interaction of AVs with automated junctions was also investigated, emphasizing the technical concerns associated with putting these systems into place.

The research of Campisi et al. [12] encourages the creation of particular, functional elements that can adapt the development of CAVs to smart cities, laying the groundwork for the defining of strategies that can be applied in the many kinds of cities of today and the coming years (for instance, by promoting educational campaigns that enable a significant decrease in the use of private vehicles and a greater focus on shared mobility, the development of induction technologies for charging electric cars and other devices, and ongoing research on the moral and social implications of using CAVs). The authors claim that road safety and environmental sustainability will significantly improve once autonomous driving technology reaches advanced levels. Future public transportation could be more effective, and traffic could be more predictable thanks to the AV revolution. This will increase the amount of open space available to city people. Additionally, there will be fewer dangers for bikers and pedestrians, who are particularly vulnerable in cities. Millions of people’s lives might be enhanced by the advantages of driverless cars and smart cities, which would also raise environmental standards. A livable city should have lots of green space, be safe, and be clean. For those with motor impairments, such as the elderly (who are becoming more numerous), those with disabilities, and marginalized groups, self-driving cars also present new chances. A livable city should have lots of green space, be safe, and be clean. For those with motor impairments, such as the elderly (who are becoming more numerous), those with disabilities, and marginalized groups, self-driving cars also present new chances. According to estimates, mobility as a service will lower a ride’s cost compared to a public transportation ticket, assisting in the dismantling of social inequality’s obstacles.

The aim of the article is to evaluate the attitude of the population to the use of autonomous vehicles.

2. Analysis of Autonomous Vehicles Possibilities and Limitations

Contemporary self-driving cars are equipped with very efficient and modern electric engines that have significant prospects for pollution reduction. The pollution generated by autonomous vehicles is reliant upon the source of their electricity. The environmental impact of Autonomous Electric Vehicles is negligible when their batteries are charged using sustainable energy sources.

AVs are much more fuel-efficient than regular cars because they use cutting-edge systems. However, they need a lot of energy to power all of their heavy-duty computers, which include cameras, radars, different sensors, and the engine itself. There is a huge need for electricity because all of the data from these different devices needs to be put together, sorted, and sent to computers as orders.

AVs may communicate with each other and the road infrastructure in intelligent methods. The directions that these cars acquire from different computers let them use the roads smarter by figuring out routes more efficiently and limiting the number of cars on the road at once. This saves fuel and lowers pollution. The last thing about AVs is that experts often say that switching to self-driving cars on a widespread basis would help reduce significantly the number of cars on the road.

For instance, families who usually have more than one car might be able to rely on a single self-driving car to meet all their needs. This is because the self-driving car could take both parents to work, pick up the kids from school, and drive back home on its own. But the main benefit of these cars is that they make it easier for people who have problems moving around. It is easy for older people to call a car without having to drive it. In this case, there are alternatives for people who have difficulties moving around to ride.

Other studies show that self-driving cars will change the way people drive in ways that are not always better for the environment. AVs, for example, permit individuals to send their vehicles on “zero-occupancy” travels. For example, rather than paying for parking, someone could send their car home while they are at work and then give another call when the day is over to come pick them up [2]. This is definitely useful, but from an environmental point of view, it may not be ideal.

Researchers also frequently suggest that people might be ready to go on longer trips because being a passenger in a car is far less stressful than operating a vehicle. AVs are also dangerous because they have a lot of technology inside them. This means that hackers have a lot of ways to hack into cars and take control of them. Another threat is that fewer people will need to drive if those cars become popular.

After doing a lot of studies, experts have found that people perceive the following ways regarding AVs:

1. The price of an AV has a big impact on people’s decisions regarding purchasing one [13]. For instance, the cost of the automated vehicle was 40,000, 50,000, or 60,000 US dollars. Subsequently, utility weights for each feature were estimated. Utility weights were then estimated for the various features. Unsurprisingly, the strongest negative utility slope was found to be the price; that is, the higher the price, the less inclined the respondents were to purchase the car [14].

2. Security factors decrease deaths caused by mistakes made by individuals [14,15]. The fact that self-driving cars might not be completely secure is one of their main issues. A driverless vehicle needs to process its surroundings to make judgment calls using perception and decision-making technology [13].

3. Policy and regulations make sure there is insurance, clear rules, and guidelines [16]. Regulation of self-driving cars, autonomous vehicles, and automated driving systems is an issue that is becoming more and more relevant in the automobile industry and is closely tied to the actual technology’s success. Several nations have enacted local laws and established guidelines for the implementation of driverless vehicles. Depending on local laws, robotaxis and self-driving trucks may also be subject to autonomous vehicle regulations [13].

4. Building the right infrastructure and integrating it with current transportation systems, like communication networks and charging points for electric vehicles [17]. A supplementary “invisible infrastructure” may also be actively used by AVs for their survival. The first of these is the ability to connect to various digital networks. Despite the common belief that no autonomous car should rely on connectivity to drive safely, practically all contemporary cars employ digital connectivity to increase their efficiency. Wider data is the second important component of invisible infrastructure, which a vehicle uses to fully understand its surroundings. Digital maps rank highest among these components. While accurate maps (preferably with centimeter-level accuracy) are currently a crucial piece of contextual information, high-level maps are necessary for all routing decisions made by the automation system to support vehicles in orientating themselves (localisation) within their environment. Other types of geographically relevant information, including traffic volumes or speed limits, also give the car helpful context or even safety-critical information. Some may also argue that a nation’s traffic laws and regulations serve as a form of infrastructure since they implicitly provide predictability and order to the road environment by outlining appropriate behavior for cars and other road users [13]. Physical road markings and traffic-control measures are, in many respects, given significance and implemented through the legal system [3,18]. In addition, policymakers are in charge of a broader range of institutional arrangements that make up the last invisible infrastructure. A vehicle must pass testing and safety regulations in order to be approved for usage on public roads. This infrastructure is also directly related to the supply of some crucial private-sector services, most notably insurance. AVs really cannot function without these additional services.

5. Social impacts of AVs contain the capacity to change transportation systems and impact society more broadly by addressing issues including energy use, air pollution, accessibility, equity, traffic congestion, rising demand for travel, effects on employment in the transportation industry, and the requirement for AV technology to be accessible to all [17]. The widespread usage of AVs will influence urban land use, private vehicle ownership perceptions, and labor demand. On the one hand, the deployment of AVs promises improved accessibility, safety, and efficiency. However, social scientists and scientists make predictions that AVs will increase socioeconomic disparity and pollution. AVs are going to present legislative and regulatory difficulties that policymakers need to be ready for. Improved safety, effective traffic control, increased accessibility, and possible environmental benefits are some of the advantages of autonomous vehicles. Nonetheless, there are still many difficulties to overcome, including public trust, technology limitations in difficult circumstances, and public confidence, which continue to be major challenges.

3. Analysis of AV Statistics on Market Size, Type, and Customers Concerns

Figure 1 shows that the world market for self-driving cars has grown by 41% in the last ten years, generating an increase in revenues from USD 147.5 billion in 2022 to USD 208.0 billion in 2023.

It is expected that the market will keep going up, and it will hit USD 282.2 billion in 2024. The revenue is expected to more than double by 2025, reaching USD 428.3 billion. Forecasts predict the market will reach USD 626.9 billion in 2026 and USD 850.6 billion in 2027. It is still growing quickly in 2027. This growth is set to accelerate further, with revenues projected to hit USD 1065.3 billion in 2028 and USD 1502.1billion in 2029. The market should make a huge amount of revenue by the end of the decade, reaching USD 2038.3 billion in 2030. It is expected that the pace will last until the beginning of 2032. It is expected that the market will bring in USD 2874 billion in 2031 and a huge USD 4206.4 billion by 2032.

According to Placek [19] research data in 2021, customers around the world were primarily concerned about safety, especially when it concerns autonomous cars: roughly 61% of respondents expressed concern about possible safety problems brought on by machine mistakes, and roughly 51% expressed concern about safety problems brought on by human error.

Evaluations of consumers’ willingness to adopt autonomous vehicles globally show that nearly fifty percent of those surveyed in 2020 said they felt comfortable traveling on driverless trains.

Additionally, customers felt reasonably at ease applying autonomous vehicles, while only about 16% of respondents said they would use an autonomous watercraft.

In Figure 2, according to Statista’s 2018 research, various countries exhibited different levels of preparedness for the integration of driverless cars, as reflected by their index scores.

The Netherlands led the rankings with a score of 27.73. Showcasing its advanced infrastructure and supportive regulatory environment for autonomous vehicles. Singapore followed closely with an index score of 26.08, highlighting its significant investments in technology and innovation. The United States scored 24.75, indicating substantial progress in autonomous vehicle readiness. Sweden and the United Kingdom demonstrated comparable levels of preparedness with scores of 24.73 and 23.99, respectively. Germany and Canada also showed strong readiness for AVs, with scores of 22.74 and 22.61. The United Arab Emirates and New Zealand scored 20.89 and 20.75, respectively, reflecting their growing efforts to support driverless car technology [21].

South Korea rounded out the top ten with an index score of 20.71, underscoring its ongoing advancements in autonomous vehicle infrastructure and policy development. These scores highlight the varying degrees of readiness among countries to adopt and integrate driverless cars into their transportation systems.

4. Countries Which Are Most Prepared for Autonomous Vehicles

The readiness of states for autonomous cars is revealed by the AVRI (Autonomous Vehicles Readiness Index) index compiled by the international audit and tax consulting services company KPMG for the third year in a row. Its results are based on 28 individual measurement units covering four different aspects of assessment: state policies and legislation defining the use and testing of autonomous vehicles; funds allocated for technological improvement and ensuring cyber security; strength, coverage, and speed of mobile communication, quality of roads, number of charging stations; consumer adaptation to new technologies.

After evaluating the most prepared countries for autonomous vehicles, the best result in the 2021 year was demonstrated by Singapore in Southeast Asia [22]. The country has recently significantly expanded the limits of autonomous car testing, which covers about 1000 km, or one-tenth of Singapore’s roads. It is desired that starting in 2022, autonomous buses will also operate in three areas of the country—about 100 bus drivers are already being retrained as security operators. The development of the network of electric car charging stations will give a big impetus to the creation of the autonomous vehicle fleet in this country. The number of these points is planned to increase from 1.6 thousand to 28 thousand over the next decade.

Two European countries are also among the top three countries most prepared for autonomous vehicles. The Netherlands, which topped the list last year, continues to lead in terms of the number of electric vehicle charging stations per capita, and in terms of road quality and market share of electric and hybrid vehicles (15%), it is in second place in both categories. In 2024, the country expanded its infrastructure for using smart devices on the roads. For example, smart traffic lights, which wirelessly send information about their status to autonomous vehicles, were installed in 60 new locations last year.

According to KPMG Baltics, due to similar weather conditions, Norway could become an example for Lithuania in the application of autonomous vehicles. It remained in third place in the AVRI index of 2021—as much as 56 percent of new cars purchased in this country last year were electric cars or hybrids. This was influenced by subsidies applied for their purchase and high taxes on vehicles with internal combustion engines and fuel.

Residents are also motivated to choose electric vehicles based on successful examples of their application. For example, in the country’s capital, Oslo, driverless buses carry passengers on special routes; electric vehicles clean the snowy and icy areas at the city airport; autonomous ferries sail along the Norwegian fjords. It is important for residents that the speed of autonomous vehicles has been increased to 20 km/h, and they are being tested even in difficult driving conditions. All of these factors increase confidence in innovations and promote development.

Case of Lithuania. In the center of Vilnius, from 2023, autonomous cars of the company “IKI Lietuva” for the delivery of goods began to operate. With this innovation, Vilnius becomes the first city in Europe to employ autonomous cars in real traffic conditions.

5. Methodology of Research

The study underwent four evaluation stages: (1) determining the criteria; (2) using a multi-criteria approach; (3) determining the criteria’s importance; and (4) analyzing the outcomes. The problem is solved using sets of multi-criteria approaches if there is a lot of information. Expert data collection techniques include surveys and scales. Experts are notified of the results of independent, anonymous polls that are frequently utilized and repeated multiple times. The Delphi technique is this approach.

It is used in situations where experts are unfamiliar with one another and do not need to consult one another in order to reach a judgment. Because some experts knew one another and answered the surveys simultaneously, this approach was dropped for the study. The Analytical Hierarchy Process, or AHP, is a decision analysis technique that combines quantitative and qualitative methodologies to solve complicated multi-objective problems [23]. Because experts utilize the AHP method to consider the relevance of criteria based on the details, giving one criterion more weight than others, this approach was rejected for the study. Several sustainable transportation options are predicted using multi-criteria analysis techniques. Additionally, it offers the chance to categorize signs from a position of expertise [24]. When conducting a study, the chosen methodology must be reliable and reflect the results of the study. In order to obtain a more accurate picture of the study, the results obtained can be compared and linked with the results obtained by different methods. Therefore, the expert ranking data evaluation and polynomial methods were chosen. Only when the expert group’s judgments are consistent and non-conflicting is the average of their opinions used to solve the problem. Expert judgment is the foundation of the majority of established and research-based techniques for determining the weights of elements. Because expert evaluations are frequently inconsistent and vary widely, the weights assigned to the criteria and their relative importance may also vary. When applying a multi-criteria evaluation method to a study, the expert group must be larger than two experts [25].

Ranking is a procedure where the most important indicator in the table is given a rating R equal to one, the second is second, and the last is m (m is the number of guidelines). Based on the multi-criteria evaluation itself, it consists of a matrix of n rows and m columns (

Table 1. Matrix of evaluation indicators (developed by the authors according to Sivilevičius [26]).

Expert Code		Indicator Marker, j = 1, 2, ..., m
		X1	X2	X3	...	Xm
i = 1, 2, ..., n	E1	Ra1	Ra2	Ra3	...	Ram
	E2	Rr1	Rr2	Rr3	...	Rrm
	E3	Rp1	Rp2	Rp3	...	Rpm
	...	...	...	...	...	...
	En	Rn1	Rn2	Rn3	...	Rnm

). Based on the methodology, a group of experts n quantitatively evaluates objects m.

The assessment can be conducted using the decimal system, percentages, fractions of units, or indicative units. The compatibility coefficient can be computed using the expert indicator ranking. If more than two experts in the expert group are involved in the study, the degree of compatibility is indicated by the compatibility coefficient. A normal distribution mean confidence interval has been computed. A qualitative feature of a measurement result that characterizes the likelihood that the measurement is true is confidence. A confidence level of 0.9 indicates that the parameter will fall within around 90% of all intervals when building particular confidence intervals numerous times. The interval’s length increases with confidence level.

For processing research data, the average of the series is calculated as follows: [27]:

$${\sum }_{i=1}^n{R}_{ij}-{\displaystyle \frac{1}{2}}n\left(m+1\right);$$

(1)

Here, m is the number of benchmarks, n is the number of experts, and R_ij is the rank of R.

$$W={\displaystyle \frac{12S}{n^{2}m\left(m^2-1\right)}}={\displaystyle \frac{12S}{n^2\left(m^3-m\right)}};$$

(2)

Here, S is the sum of the squares of the deviation from the arithmetic mean, and W is the concordance coefficient. By Formula (3), the Pearson criterion χ² is calculated as follows:

$${\chi }^2=n\left(m-1\right)W={\displaystyle \frac{12S}{nm\left(m+1\right)}}.$$

(3)

The concordance coefficient W_min is calculated as follows:

$$W_{\mathrm{m}\mathrm{i}\mathrm{n}}={\displaystyle \frac{{\chi }_{v,\alpha }^2}{n\left(m-1\right)}};$$

(4)

$$S_{\mathrm{m}\mathrm{a}\mathrm{x}}={\displaystyle \frac{n^{2}m(m^2-1)}{12}};$$

(5)

$$\overline{R}={\displaystyle \frac{1}{2}}n\left(m+1\right).$$

(6)

The Kendall ranks’ conformity coefficient is calculated in accordance with the experts’ evaluation indicators (6), which are used to determine the consistency of their opinions. The Formula (4) is used to calculate the lowest value of the concordance coefficient in W_min. The Pearson criteria can be used to determine the significance of the concordance coefficient and to establish a threshold value for the concordance coefficient, which allows for the coordination of expert assessments. In the absence of associated ranks, the correlation coefficient is defined as the ratio of the obtained S to the maximum S_max according to Formula (5), where S calculated according to Formula (1) is the true sum of squares. Regression analysis is useful in research as a tool to examine the relationship between two variables. Ordinal regression models are used to determine the level of adoption of ordinal variables that capture the degree of novelty or complexity of technology use. Polynomial choice models are useful when there are three or more values to choose from [28].

6. Research

It takes expertise and professionals from a variety of sectors to identify the obstacles and possibilities for the adoption of the investigated AV. The expertise and experience of the experts who took part in the study to evaluate the elements determine how reliable the rankings of the overall expert group are. Finding the barriers and opportunities for the adoption of the studied AV requires knowledge and professionals from a range of industries. The reliability of the rankings of the entire expert group depends on the knowledge and experience of the experts who participated in the study in assessing the aspects. Among the specialists are two land transport scientists, practitioners, and college students, two practicum experts in transport engineering, and four practical experts in logistics. Only when all experts’ solutions are consistent can expert assessment be used as a solution to the issue.

The research was carried out in 2025, January–February. Eight experts (E8, E7, E6, E5, E4, E3, E2, E1) with university education—MASTER’S degree and 10 years of continuous professional experience—participated in the study (

Table 2. Expert data (developed by authors).

Expert	Education	Professional Experience	Occupation
E1	Master	10	manager
E2	Master	11	manager
E3	Master	12	manager
E4	Master	11	manager
E5	Doctor	15	manager
E6	Doctor	20	manager
E7	Master	18	manager
E8	Master	15	manager

).

With knowledge of each expert’s qualifications and real-world experience, the writers put together a group of specialists. The chosen specialists decided to use established definitions to score the parameters listed in the questionnaire. The reasons why Lithuanian transport companies have yet to adopt autonomous cars to transport passengers or freight to improve service quality were ranked by experts on a scale of 1 (extremely important) to 8 (not at all important). The following parameters were assessed:

• AA1—the cost of autonomous cars;
• AA2—lack of security;
• AA3—there are no clear rules for insurance of such cars and no ready-made legal basis for operation;
• AA4—there is no proper infrastructure;
• AA5—unsecured digital connection required to transmit large amounts of data for AV control;
• AA6—current traffic and speed restrictions;
• AA7—lack of testing and safety regime standards;
• AA8—autonomous cars will increase pollution and economic inequality.

Expert rank distribution is provided in Figure 3.

The analysis and calculation data for the eight expert criteria (order of importance: 1 as the most important, 8 as the least important) are presented in

Table 3. Data of the Obtained Ranks (developed by authors).

Formula	Criterion Encryption Symbol (m = 8)
	AA1	AA2	AA3	AA4	AA5	AA6	AA7	AA8
${\sum }_{\mathit{i}=1}^{\mathit{n}}{R}_{ij}$	50	45	40	38	37	35	23	20
${\overline{R}}_j={\displaystyle \frac{{\sum }_{i=1}^n{R}_{ij}}{n}}$	6.25	5.63	5.00	4.75	4.63	4.38	2.88	2.50
${\sum }_{i=1}^n{R}_{ij}-{\displaystyle \frac{1}{2}}n\left(m+1\right)$	14	9	4	2	1	−1	−13	−16
${\left[{\sum }_{i=1}^n{R}_{ij}-{\displaystyle \frac{1}{2}}n\left(m+1\right)\right]}^2$	196	81	16	4	1	1	169	256

.

Having the rank results, we calculate the concordance coefficient according to Equation (7). It should be noted that when calculating the concordance coefficient according to Equation (7), there are no associated ranks:

$$W={\displaystyle \frac{12S}{n^2\left(m^3-m\right)}}={\displaystyle \frac{12·724}{8^2\left(8^3-8\right)}}=0.2693.$$

(7)

The importance of not using autonomous cars in transportation, number m > 8. To calculate the random variable, the concordance coefficient is calculated according to Equation (8):

$${\chi }^2=n\left(m-1\right)W={\displaystyle \frac{12S}{nm\left(m+1\right)}}={\displaystyle \frac{12·724}{8·8\left(8+1\right)}}=15.083.$$

(8)

The average ratings indicate what the experts generally believe, and the opinions of the experts who replied are considered consistent because the calculated value of 15.083 is higher than the critical value, which is equivalent to 14.0671. Equation (9) is used to get the concordance coefficient W_min’s lowest value:

$$W_{\mathrm{m}\mathrm{i}\mathrm{n}}={\displaystyle \frac{{\chi }_{v,\alpha }^2}{n\left(m-1\right)}}={\displaystyle \frac{14.0671}{8\left(8-1\right)}}=0.2512<0.2693.$$

(9)

The opinions of all eight respondents regarding the eight criteria for not using autonomous cars to transport people or freight in Lithuanian transport companies to improve the quality of service are significant and are still regarded as agreeable if W_min = 0.2512 < 0.2693. The calculation of the confidence interval is performed. Obtained confidence interval: [1.61; 5.35].

The impact of autonomous cars, which is measured as Q_j, is significant for enhancing the freight and passenger transportation of Lithuanian transport businesses. The collected data are significant and can be seen in

Table 4. Results of Q_j (compiled by the authors).

Indicator Marker	Criterion Encryption Symbol (m = 8)
	AA1	AA2	AA3	AA4	AA5	AA6	AA7	AA8	Sum
qj	0.1736	0.1563	0.1389	0.1319	0.1285	0.1215	0.0799	0.0694	1
dj	0.8264	0.8438	0.8611	0.8681	0.8715	0.8785	0.9201	0.9306	7
Qj	0.1181	0.1205	0.1230	0.1240	0.1245	0.1255	0.1314	0.1329	1
Qj′	0.0764	0.0938	0.1111	0.1181	0.1215	0.1285	0.1701	0.1806	1
Distribution of importance of criteria	8	7	6	5	4	3	2	1	–

in order of importance.

Based on expert review and calculations, it has been established that indications of Lithuanian transport companies’ not using autonomous cars to improve service quality are considerable. The components and standards ranked by significance are as follows:

(1)

AA8—autonomous cars will increase pollution and economic inequality;

(2)

AA7—lack of testing and safety regime standards;

(3)

AA6—current traffic and speed restrictions;

(4)

AA5—unsecured digital connection required to transmit large amounts of data for AV control;

(5)

AA4—there is no proper infrastructure;

(6)

AA3—there are no clear rules for insurance of such cars and no ready-made legal basis for operation;

(7)

AA2—lack of security;

(8)

AA1—the cost of autonomous cars.

The impact of autonomous vehicle use on the quality of passenger or freight transportation services is presented in Figure 4.

The estimates of the second-degree polynomial criteria (Figure 4) show that a 2-fold increase in the use of autonomous vehicles in transportation services resulted in a 1.2-fold improvement in the quality of transportation services. The analysis uses polynomial curve fitting. The closer the regression coefficient R² is to unity, the better the regression curve fits the experimental data. The closer the answer is to 1, the stronger the correlation between the variables under study.

In order to improve the quality of services, experts were asked to evaluate how their company’s operations could be affected if they began using autonomous automobiles. The following standards were evaluated:

• AV1—round-the-clock mobility is ensured;
• AV2—reduced number of cars on the road and reduced CO₂;
• AV3—reduced number of accidents;
• AV4—the characteristics of autonomous car control would increase (more attention to research);
• AV5—positive impact on the environment would increase;
• AV6—increasing user accessibility;
• AV7—scenarios of increasingly complex and intractable processes.

8 expert rank distribution is provided in Figure 5.

The analysis and calculation data for the eight expert criteria (order of importance: 1 as the most important, 8 as the least important) are presented in

Table 5. Data of the Obtained Ranks (developed by authors).

Formula	Criterion Encryption Symbol (m = 7)
	AV1	AV2	AV3	AV4	AV5	AV6	AV7
${\sum }_{i=1}^n{R}_{ij}$	8	16	28	40	34	42	56
${\overline{R}}_j={\displaystyle \frac{{\sum }_{i=1}^n{R}_{ij}}{n}}$	1.00	2.00	3.50	5.00	4.25	5.25	7.00
${\sum }_{i=1}^n{R}_{ij}-{\displaystyle \frac{1}{2}}n\left(m+1\right)$	−24	−16	−4	8	2	10	24
${\left[{\sum }_{i=1}^n{R}_{ij}-{\displaystyle \frac{1}{2}}n\left(m+1\right)\right]}^2$	576	256	16	64	4	100	576

.

It should be noted, as follows, that when calculating the concordance coefficient according to Equation (10), there are no associated ranks:

$$W={\displaystyle \frac{12S}{n^2\left(m^3-m\right)}}={\displaystyle \frac{12·1592}{8^2\left(7^3-7\right)}}=0.8884.$$

(10)

The significance of how their business operations would be affected if they began using autonomous vehicles for service, with a number m > 7. After that, a random variable is obtained by calculating the concordance coefficient using Equation (11) as follows:

$${\chi }^2=n\left(m-1\right)W={\displaystyle \frac{12S}{nm\left(m+1\right)}}={\displaystyle \frac{12·1592}{8·7\left(7+1\right)}}=42.6429.$$

(11)

The calculated value of${\chi }^2$ = 42.6429, is greater than the critical value (equal to 12.5916). Therefore, the opinions of the responding experts are perceived as consistent, and the average ratings indicate the general opinion of the experts.

Equation (12) is used to get the concordance coefficient W_min’s lowest value as follows:

$$W_{\mathrm{m}\mathrm{i}\mathrm{n}}={\displaystyle \frac{{\chi }_{v,\alpha }^2}{n\left(m-1\right)}}={\displaystyle \frac{12.5916}{8\left(7-1\right)}}=0.2623<0.8884.$$

(12)

The opinions of all eight respondents on the seven criteria about how their company’s operations would change if they began using autonomous automobiles to improve service quality are significant and are still regarded as agreeable if W_min = 0.2623 < 0.8884. The calculation of the confidence interval is performed. Obtained confidence interval: [2.50; 4].

The impact of autonomous vehicles is calculated as Q_j. The data obtained are significant and are presented in

Table 6. Results of Q_j (compiled by the authors).

Indicator Marker	Criterion Encryption Symbol (m = 7)
	AV1	AV2	AV3	AV4	AV5	AV6	AV7
qj	0.0357	0.0714	0.1250	0.1786	0.1518	0.1875	0.2500
dj	0.9643	0.9286	0.8750	0.8214	0.8482	0.8125	0.7500
Qj	0.1607	0.1548	0.1458	0.1369	0.1414	0.1354	0.1250
Qj′	0.2500	0.2143	0.1607	0.1071	0.1339	0.0982	0.0357
Distribution of importance of criteria	1	2	3	5	4	6	7

.

The following are the elements and criteria assessed by the level of importance:

(1)

AV1—round-the-clock mobility is ensured;

(2)

AV2—reduced number of cars on the road and reduced CO₂;

(3)

AV3—reduced number of accidents;

(4)

AV4—the characteristics of autonomous car control would increase (more attention to research);

(5)

AV5—positive impact on the environment would increase;

(6)

AV6—increasing user accessibility;

(7)

AV7—scenarios of increasingly complex and intractable processes.

Figure 6 shows the impact of autonomous vehicle use on the decreasing number of cars on the road.

A six-fold increase in the use of autonomous cars would lead to a four-fold drop in the number of vehicles on the road, according to estimations based on the second-degree polynomial criterion (Figure 6). The regression coefficient R² shows a strong correlation between the indicators under study.

Experts put the criteria that outline the requirements for using autonomous vehicles in Lithuania to the test. The following characteristics were assessed:

• AT1—lack of state policy and legislation;
• AT2—lack of legislation for testing and using autonomous cars;
• AT3—there is a lack of funds for technological development in the country;
• AT4—lack of funds for cybersecurity;
• AT5—lack of (too weak) mobile connection strength, speed;
• AT6—lack of quality of roads (infrastructure);
• AT7—lack of charging stations;
• AT8—lack of willingness and consumer adaptation to new technologies.

8 expert rank distribution is provided in Figure 7.

The analysis and calculation data for the eight expert criteria (order of importance: 1 as the most important, 8 as the least important) are presented in

Table 7. Data of the Obtained Ranks (developed by authors).

Formula	Criterion Encryption Symbol (m = 8)
	AT1	AT2	AT3	AT4	AT5	AT6	AT7	AT8
${\sum }_{i=1}^n{R}_{ij}$	25	19	21	28	44	40	47	64
${\overline{R}}_j={\displaystyle \frac{{\sum }_{i=1}^n{R}_{ij}}{n}}$	3.13	2.38	2.63	3.50	5.50	5.00	5.88	8.00
${\sum }_{i=1}^n{R}_{ij}-{\displaystyle \frac{1}{2}}n\left(m+1\right)$	−11	−17	−15	−8	8	4	11	28
${\left[{\sum }_{i=1}^n{R}_{ij}-{\displaystyle \frac{1}{2}}n\left(m+1\right)\right]}^2$	121	289	225	64	64	16	121	784

.

It should be noted, as follows, that when calculating the concordance coefficient according to Equation (13), there are no associated ranks:

$$W={\displaystyle \frac{12S}{n^2\left(m^3-m\right)}}={\displaystyle \frac{12·1684}{8^2\left(8^3-8\right)}}=0.6265.$$

(13)

The importance of the criteria, which shows what is needed to be able to use autonomous cars in Lithuania, is number m > 8. To calculate the random variable, the concordance coefficient is calculated according to Equation (14) as follows:

$${\chi }^2=n\left(m-1\right)W={\displaystyle \frac{12S}{nm\left(m+1\right)}}={\displaystyle \frac{12·1684}{8·8\left(8+1\right)}}=35.083.$$

(14)

The calculated value of${\chi }^2$ = 35.083, is greater than the critical value (equal to 14.0671). Therefore, the opinions of the responding experts are perceived as consistent, and the average ratings indicate the general opinion of the experts.

Equation (15) is used to get the concordance coefficient W_min’s lowest value as follows:

$$W_{\mathrm{m}\mathrm{i}\mathrm{n}}={\displaystyle \frac{{\chi }_{v,\alpha }^2}{n\left(m-1\right)}}={\displaystyle \frac{14.0671}{8\left(8-1\right)}}=0.2512<0.6265.$$

(15)

All eight respondents’ perspectives on the eight criteria that outline the requirements for using autonomous automobiles in Lithuania are significant and are still regarded as agreeable if W_min = 0.2512 < 0.6265. The calculation of the confidence interval is performed. Obtained confidence interval [3.20; 1.90].

The impact of the criteria shows what is needed to be able to use autonomous cars in Lithuania.is calculated as Q_j. The data obtained are significant and are presented in

Table 8. Results of Q_j (compiled by the Authors).

Indicator Marker	Criterion Encryption Symbol (m = 8)
	AT1	AT2	AT3	AT4	AT5	AT6	AT7	AT8	Sum
qj	0.0868	0.0660	0.0729	0.0972	0.1528	0.1389	0.1632	0.2222	1
dj	0.9132	0.9340	0.9271	0.9028	0.8472	0.8611	0.8368	0.7778	7
Qj	0.1305	0.1334	0.1324	0.1290	0.1210	0.1230	0.1195	0.1111	1
Qj′	0.1632	0.1840	0.1771	0.1528	0.0972	0.1111	0.0868	0.0278	1
Distribution of importance of criteria	3	1	2	4	6	5	7	8	–

.

The following are the elements and criteria assessed by level of importance:

(1)

AT2—lack of legislation for testing and using autonomous cars;

(2)

AT3—there is a lack of funds for technological development in the country;

(3)

AT1—lack of state policy and legislation;

(4)

AT4—lack of funds for cybersecurity;

(5)

AT6—lack of quality of roads (infrastructure);

(6)

AT5—lack of (too weak) mobile connection strength, speed;

(7)

AT7—lack of charging stations;

(8)

AT8—lack of willingness and consumer adaptation to new technologies.

The impact of the dependence on autonomous cars on public investment and favorable policy decisions is presented in Figure 8.

According to the projections of the second-degree polynomial criteria (Figure 8), the use of autonomous vehicles would increase by 1.7 if the state doubled its investments in this technology and made favorable legislative decisions. The regression coefficient shows a strong correlation between the indicators under study.

7. Discussion

Our study’s views regarding the practical application of autonomous vehicles highlight the necessity of acknowledging its critical role in the sustainable transition to a green economy. The research was conducted regarding the specifics of Lithuanian transport enterprises. Also, the experts were selected from limited organizations. In the future, it would be useful to carry out research on a wider scale geographically, attracting more diverse experts.

The article highlights the necessity of ongoing advancements in the creation of autonomous vehicle rules and infrastructure. Residents of many nations differ in their willingness to embrace and incorporate autonomous automobiles into their transportation networks.

Residents are also encouraged to pick electric automobiles by the effective acceptance of these vehicles. Even in challenging driving circumstances, it is crucial for locals that autonomous vehicles accelerate. AVs are thought to be safer and more effective vehicles for traffic control. Because of the methods they employ, these vehicles are more cost-effective than traditional ones. The study’s other finding indicates that autonomous vehicles would alter driving habits, which may or may not be better for society as a whole.

8. Limitation

Only professionals employed by the biggest logistics and transportation firms in Lithuania participated in this study. The obtained expert opinions are subjective because experts from different areas were not interviewed. Numerous logistics firms are working to advance AV technologies and plan to use them in the future.

New AV technology development is costly and could restrict customer accessibility. There has been no evaluation of AV’s impact on how the transportation process is organized. It is also necessary to look into synergy with public transportation systems.

As the study is based on expert opinions without empirical validation, in the future, it would be useful to verify, for example, the impact on service quality and CO₂ reduction by real audiovisual system tests or simulation data. The omission of scenario analyses, such as varying audiovisual penetration levels or sensitivity studies, further limits the practical applicability of the findings to policymakers. Therefore, the analysis of the abovementioned aspects must be as comprehensive as possible.

Besides this, in the future, the research on recent developments in autonomous driving should be reviewed to ensure their completeness.

9. Conclusions

• Autonomous vehicles exhibit considerable potential by facilitating more efficient driving through lighter and safer automobiles while encouraging shared and on-demand mobility to exemplify possibilities in the future [29]. The extensive use of self-driving cars would lead to a decrease in the quantity of cars on the roads, hence reducing CO₂ emissions and road fatalities.
• After doing a lot of studies, experts have found that the following factors greatly impact people’s decisions regarding purchasing an AV: price, safety, regulation policy, social indicators, and building the right infrastructure and integrating it with current transportation systems.
• In this article’s evaluation of the most prepared countries for autonomous vehicles, the best result in 2021 was demonstrated by Singapore in Southeast Asia [18]. In the European countries, the Netherlands, which topped the list for 2023, continues to lead in terms of the number of electric vehicle charging stations per capita and in terms of road quality and market share of electric and hybrid vehicles (15%); it is in second place in both categories.
• In this article, the authors’ evaluation was based on the multi-criteria evaluation method; a group of eight experts evaluated the research objects by ranking them. Based on expert review and calculations, it has been established that indications of Lithuanian transport companies’ not using autonomous cars to improve service quality are considerable as follows: (1) AV will increase pollution and economic inequality; (2) the lack of testing and safety regime standards; (3) current traffic and speed restrictions; (4) unsecured digital connection; (5) there is no proper infrastructure; (6) there are no clear rules for insurance of such cars and no ready-made legal basis for operation; (7) lack of security; (8) the cost of autonomous cars.
• The impact of autonomous vehicles on the quality of service was calculated by level of importance: (1) round-the-clock mobility is ensured; (2) reduced number of cars on the road and reduced CO₂; (3) reduced number of accidents; (4) the characteristics of autonomous car control would increase; (5) positive impact on the environment would increase; (6) increasing user accessibility.
• As audiovisual technologies continue to develop, important research on recent advancements in autonomous driving—such as the impact of autonomous vehicles on service quality and CO₂ reduction—could be conducted using real audiovisual system tests or simulation data. This would enhance the practical applicability of the findings for policymakers.