Autonomous Vehicles as Viewed by Future Users in Poland

Abstract

The article discusses the issue of how potential users perceive automated and autonomous vehicle (AV) technology, emphasizing that its sophistication and reliability are crucial for the safety of driving vehicles with a level of automation starting from the third degree. Safety plays a significant role in determining public acceptance of autonomous vehicles. Without the acceptance of potential users and their willingness to use smart electric vehicles, it is challenging to talk about the rapid implementation of autonomous vehicles, despite their undeniable benefits. The article presents an analysis of survey results regarding public opinion in Poland on this matter. The survey utilized various methods, including CAWI surveys, Focus Group Interviews (FGI), and Individual In-depth Interviews (IDI). The CAWI survey involved 1067 respondents in Poland, with a 95% confidence level and a maximum margin of error of 3%, representing a representative sample of the Polish population aged 16 and above. The article examines issues that illustrate respondents’ acceptance levels concerning perceived security and trust in AVs. The obtained results are compared with opinions from surveys conducted in other countries. Additionally, the article attempts to identify reasons for the low assessment of the technologies among respondents and proposes measures to boost confidence in automated and autonomous vehicles.

1. Introduction

The turn of the 20th and 21st centuries has witnessed a re-evaluation of many concepts and the emergence of new priorities in socio-economic development. This includes the shift towards sustainable development based on three pillars: ecology, economics, and a comprehensive consideration of human needs. Society has evolved from an industrial society to an information society, and now, it is moving towards a knowledge society. This evolution spans various dimensions, including social, economic, scientific, and technological aspects. As a result, new areas of knowledge are emerging, accompanied by innovative technological ideas that are quickly transitioning into implementation. An intriguing recent development is the integration of artificial intelligence techniques into a crucial domain of knowledge and a vital component of the contemporary economy and human life: transportation. Transportation holds a pivotal position in the modern world. It aspires to be environmentally friendly, energy and resource-efficient, and intelligent. Additionally, it should cater to individuals with limited mobility. Presently, the most significant challenge in the realms of information technology and automotive industries is achieving autonomy in transportation. This entails advancing technology to a level where it not only assists drivers but also enables the substitution of a human driver with a “machine” that employs artificial intelligence [1].

Harnessing advancements in artificial intelligence allows for a completely new level of collaboration among road users, offering substantial benefits to both individuals and the overall mobility system. These benefits encompass enhanced safety, accessibility, and sustainability. Nonetheless, any novel solution encounters various barriers during its implementation, one of which is the level of user acceptance, or more accurately, the lack thereof. This lack of acceptance stems from multiple reasons, including technology that is not yet flawless and a deficiency in knowledge among potential users. Therefore, it is essential to contemplate the extent of knowledge and how it should be conveyed to the public to instill confidence in sensor technologies, which constitute a crucial component of automated vehicles and autonomous AVs (Automated Vehicles). This article draws upon public opinion surveys conducted as part of the AV-PL-ROAD project, titled “Polish Road to Automation of Road Transport”. The issue of AVs’ deployment is on the strategic agenda of the Polish government, which underscores the significance of any action that can boost user trust in this new technology, and understanding the reasons for its lack thereof is extremely important. The terms “automated vehicle” and “autonomous vehicle” are often used interchangeably. It is important to establish the object of analysis from the outset. As a result of the AV-PL-ROAD project, the classification [2] discussed by the U.S. National Highway Traffic Safety Administration (NHTSA), the U.S. Society of Automotive Engineers (SAE), and other industry associations includes the following levels:

• Level 0 (no driving automation)—manual control;
• Level 1 (driver assistance)—a single automated driver assistance system;
• Level 2 (partial driving automation)—the presence of an advanced driver assistance system or ADAS (Advanced Driver Assistance Systems);
• Level 3 (conditional driving automation)—the ability to monitor the environment and the vehicle’s capacity to make autonomous decisions;
• Level 4 (high level of driving autonomy)—vehicles can intervene autonomously (with geographical location restrictions);
• Level 5 (fully autonomous car)—vehicles do not require human attention (no geographical restrictions).

Levels 3 to 5 are of particular concern to users because, with increasing levels of automation, sensory measurement systems play an increasingly crucial role in identifying objects and their characteristics in the vicinity of the vehicle [1].

The autonomous vehicle market in Poland is currently in the early stages of development, offering significant opportunities for Polish companies in the automotive, IT, and related industries. Polish firms have the chance to become suppliers of essential components for autonomous vehicles. One example is the startup Robotec.ai, which develops monitoring and communication solutions for autonomous vehicles, including lidars, biometric sensors to monitor the driver’s state of concentration, and ADAS systems. Software plays a crucial role in autonomous vehicles, and numerous Polish companies and startups are involved in designing software for recognizing road signs and infrastructure elements, creating vehicle environment maps, and determining vehicle location using SLAM methodologies and trajectory control algorithms.

The electrification of vehicles is closely tied to vehicle autonomy, and Polish companies producing batteries for electric vehicles, such as Impact, contribute to the development of the autonomous vehicle market. Concepts and prototypes of autonomous PRT (Personal Rapid Transit) vehicles have also emerged.

However, the production of autonomous vehicles is just one aspect of the challenge of introducing them to Polish roads. Legal regulations and infrastructure are two other critical factors. Currently, autonomous vehicles are not legal in Poland, and testing is only allowed for research purposes with permission from the traffic management authority and the use of a designated test track.

The introduction of autonomous vehicles in Poland should be preceded by the expansion of existing road infrastructure to incorporate Vehicle-to-Infrastructure (V2I) communication elements. Poland is currently not considered in the ranking of autonomous vehicle readiness, but the AV-PL-ROAD project has significantly contributed to research on legislative changes needed for the safe implementation of road automation in Poland. This includes international context, public acceptance of autonomous vehicles, and an assessment of the country’s infrastructure readiness.

As a result of these efforts, the Autonomous and Connected Vehicle Competence Center was established at the Institute of Automotive Transport. The center serves as an expert institution supporting government efforts, particularly those of the Ministry of Infrastructure, in the field of autonomous and connected vehicles.

Ongoing projects, initiated by the Ministry of Infrastructure and funded under the GOSPOSTRATEG Program, aim to create a database of road scenarios for testing perception systems for autonomous vehicles under conditions typical for Polish road infrastructure.

It is projected that autonomous vehicles will hit Polish roads around 2030. However, several challenges need to be addressed before autonomy becomes a reality, including the construction of intelligent and costly road infrastructure, continuous technological advancements, necessary changes in the Polish legal system, and societal attitudes. Public opinion is crucial in the introduction of autonomous vehicles in Poland, as positive sentiment toward new technologies can significantly accelerate the adaptation process. Therefore, it is essential to consider and study public opinion during the planning and implementation of such innovations and work toward increasing societal acceptance in case of low or no approval, as outlined by the authors of this article.

The successful implementation of innovative technologies, including automated and autonomous driving technology, heavily relies on the attitudes and expectations of future users and the general public. The transition from conventional to automated driving represents significant changes for all road users. New solutions invariably raise concerns about their effectiveness and safety. Building trust in these solutions is a long-term process, and researchers worldwide employ questionnaires and surveys to assess current acceptance levels across various contexts. Studies have explored the acceptance of AVs in different countries [3,4,5] and for different applications, such as privately owned automated vehicles [6] versus their use in public transportation [7,8]. Research is conducted in simulators [9] and real-world settings [10] to understand how exposure to AVs affects acceptance rates.

As part of the Gospostrateg project—PL-AV-ROAD—“Poland’s Road to Autonomous Transportation”, survey research was conducted on a representative sample of the Polish population aged 16 and above. The aim of the research was to determine consumers’ readiness to use autonomous vehicles and understand their expectations regarding these vehicles.

The authors formulated the following hypothesis: Consumer readiness is closely linked to the level of acceptance of new technologies, which, in turn, increases as knowledge in this area expands, and with the opportunity to interact with real automated vehicles.

Additionally, the authors put forward the thesis that coordinated actions aimed at increasing societal awareness and shaping a positive perception through various initiatives will significantly contribute to a favorable public opinion towards autonomous vehicle technology. Consequently, the manuscript attempts to understand the social perspective of introducing autonomous vehicles. The insights gained from these studies can contribute to a better understanding of the challenges and barriers associated with the social adaptation of this new technology in Poland. Furthermore, a research approach based on survey studies on a representative sample of the population can provide valuable data on the attitudes and expectations of society towards autonomous vehicles.

The research results obtained can serve as a valuable source of information for decision-makers, entrepreneurs, and other stakeholders involved in the process of adapting this new technology. It can also shed light on the perception of the Polish autonomous vehicle market globally and among the Polish society.

2. Research Method

Public opinion surveys regarding the acceptance of road transport autonomization technologies were commissioned by the Ministry of Infrastructure and conducted as part of the GOSPOSTRATEG program. The research was carried out by the Economics Department of the Road and Bridge Research Institute and the Research and Development Center of the University of Economics in Katowice. Surveys were prepared for both a representative group of respondents from the Polish population and experts from the TSL (Transport Forwarding Logistics) industry by research teams from the Ministry of Infrastructure and the Warsaw University of Technology [11].

A population study of individuals aged 16 and older was conducted using a randomly stratified sample with quota variables: gender, age (six options), size of the place of residence (five options), and all provinces in Poland. Within each stratum, a simple random sampling of respondents was performed. In the first step, the sampling frame was narrowed down to individuals meeting the main research criterion, i.e., respondents aged 16 and older. Based on data from the official registers of the Central Statistical Office, a representative sample structure was determined based on the combined distribution of gender, age category, and place of residence, as well as marginally with respect to the province. Subsequently, independently for each target cell of the sample structure (strata), the target size to be achieved in the implementation was determined. In the next step, the size of each stratum was divided by the expected response rate for a given category, thus establishing the minimum number of required invitations that were sent to achieve the required implementation threshold. The study was conducted in November 2021.

The second research group consisted of representatives of companies in the TSL (Transport, Shipping, and Logistics) industry, including representatives of transport companies, within the range of 70–80%. The research was conducted on a purposefully selected sample of 155 respondents, ensuring that participants in the study were individuals who, due to their characteristics, should undergo such research, i.e., managers of TSL companies, managers of TSL industry associations, drivers, dispatchers, and freight forwarders.

The survey employed the following methods: CAWI surveys (Computer Assisted Web Interview), FGI interviews (Focus Group Interview), and IDI interviews (Individual In-depth Interview).

2.1. CAWI Survey Structure

The CAWI survey covered 1067 respondents in Poland, with a confidence level of 95% and a maximum margin of error of 3%, making it a representative sample of the Polish population aged 16 and above. Within this group, 52.2% were women. The representative sample’s structure was determined based on data from official GUS registers, considering the overall distribution of gender, age categories, place of residence, and region (16 territorial units). The survey consisted of 68 thematically grouped questions. Information regarding the research issues and questions is presented in

Table 1. Information about research issues and variables. The columns are labeled as follows: “Group”—questions related to one research issue (group variable), “Label”—description of the issue that answers are meant to explain, “Position”—question number within the defined group, “Variable Name”—the name of the variable related to the specific content of the question (subordinate variable components).

Group	Label	Position	Variable Name
Id	Interview ID	1
Voivodeship	Region of residence	2
P1	Introduction	3
P2_1–P2_2	Age category	4, (5-)
P3	Gender	6
P4	Education level	7
P5	Type of education	8
P6_1–P6_2	Size of the lacality	9, (10-)
P7	Type of vehicle driven	11
P8	Use of private transportation means	12
P9_1–P9_7	Assistance systems in your vehicle	13, 14, 15, 16, 17, 18, 19	Parking sensors (1), cruise control (2), adaptive cruise control (3), lane-keeping assistant (4), lane change assistant (5), parking assistant (6), no system (7)
P10	[Label missing]	-	-
P11	Level of knowledge about autonomous vehicles	20
P12	Level of knowledge about autonomous vehicle testing worldwide	21
P13	Level of knowledge about autonomous vehicle testing in Poland	22
P14	Evaluation of the informative value of mass media	23
P15	General opinion about autonomous vehicles	24
P16_1–P16_8	Anticipated benefits of implementing and using autonomous vehicles	25, 26, 27, 28, 29, 30, 31, 32	Enhancing safety (1), comfort (2), transportation accessibility (3), traffic reduction (4), shorter travel times (5), emission reduction (6), job creation (7), economic growth (8)
P17_1–P_17_4	Acceptance of autonomous vehicle-based transportation	33, 34, 35, 36	New knowledge (1), media information (2), personal experience (3), all of the above conditions (4)
P18_1–P_18_6	Reasons for the lack of universal acceptance of autonomous vehicles	37, 38, 39, 40, 41, 42	Technology reliability (1), fear (2), reluctance to give up control (3), inefficiency (4), high cost (5), all of the above conditions (6)
P19_1–P_19_8	Barriers to the introduction of technology	43, 44, 45, 46, 47, 48, 49, 50	Lack of accident liability (1), absence of traffic laws (2), low level of technology (3), low safety (4), risk of data loss (5), poor infrastructure (6), lack of mixed traffic capability (7), absence of professional support (8)
P20_1–P20_4	Scenarios of traffic involving autonomous vehicles	51, 52, 53, 54	Without passengers (1), freight transport (2), buses (3), taxis (4)
P21_1–P21_3	Willingness to purchase an autonomous car compared to the price of a traditional car	55, 56, 57	Exceeds (1), equals (2), is less than (3
P22_1–P22_6	Use of time while driving	58, 59, 60, 61, 62	Work (1), sleep (2), entertainment (3), vehicle control (4), other (5))
		63	Describe other
P23_1–P23_5	Preferred mode of using a driverless car	64, 65, 66, 67	Private (1), on-demand (2), public (3), other (4)
		68	Describe other

. There were 23 general research issues defined in the “Group Label” column. The test question numbers related to each general issue are listed in the “Position” column. The “Variable Name” column specifies labels that can be assigned to the content of specific questions within a given research group. An empty field in this column indicates that only one question was asked in the group, which in this case serves as the group label. Detailed question content can be found on the website of the Ministry of Infrastructure or obtained through access to public information channels.

2.2. Variable Characteristics

In accordance with the CAWI research methodology, the survey questionnaire included closed-ended questions utilizing three types of scales: nominal with responses such as “yes/no”, ordinal based on a cafeteria of single or multiple-choice responses, and absolute scales indicating direct numerical values. The nominal scale was used in groups such as Voivodeship, P3, P4, P5, P7, and P8. These data essentially serve as the questionnaire’s metrics. The ordinal scale was primarily used for variables described by responses to the survey questions and pertains to groups P9–P23. The absolute scale was employed in groups P2 and P6 (the questionnaire’s metrics). For responses classified using the ordinal scale, the range of permissible statistical operations is limited to non-parametric analyses. However, if the responses are expressed using a Likert scale (responses on a scale of 1–5 or 1–7), the data from measurements can be treated as quantitative (scalar). This situation also occurs when analyzing group variables (P16–P23), which consist of combinations of multiple subordinate questions with dichotomous “yes/no” responses of identical weight value. In this case, group variables automatically become Likert-scale data and are considered scalar. If their distributions are also normal, they can be analyzed parametrically.

For data related to Voivodeship, P3, P4, P5, and P7, we have nominal data representing categorical characteristics without a natural hierarchy. These should be analyzed non-parametrically. However, it is also possible to propose converting categories into a numerical scale, thereby giving these variables a scalar nature. This would enable the examination of their relationship with other variables described through survey responses. A more complex situation arises with data regarding responses to open-ended questions, specifically questions number 63 and 68. In these data, it is observed that some responses are repetitive. By defining unique response categories (distinct values), they can form the basis for creating an individual nominal scale (here, a minimum number of elements in each category could also be imposed). However, the number of unique categories obtained is quite large, making it impractical to introduce numerical equivalents. Therefore, in the subsequent analysis, subordinate variables related to these questions will not be considered. To summarize these considerations, it can be stated that most group variables can be regarded as scalar because they are either Likert-scale data or nominal data that can be transformed into numerical characteristics. However, there are two subordinate variables for questions 63 and 68 that cannot be treated as scalar and will not be included in the survey analysis.

2.3. Research Methodology

In this study, we will aim to achieve our research objectives by analyzing the relationships between group variables described by scalar data. We have at our disposal both parametric and non-parametric procedures. In non-parametric procedures, we can perform operations such as ranking or non-parametric tests (median, Spearman’s rank correlation coefficient, Kendall’s tau-b, Wilcoxon’s test, and Friedman’s test). The resulting measures are qualitative in nature, indicating relationships like greater or lesser, or whether the variables come from the same distribution. An example of using this procedure to analyze test results is presented in a work by Choromański, Grabarek, and Kozłowski in Transport Research Part F [12]. For parametric procedures (mean value, Pearson’s correlations for normally distributed data, ANOVA tests), we can obtain a more precise description of the strength of relationships in the form of numerical values, such as the Pearson correlation coefficient. Parametric procedures provide better qualitative and quantitative results regarding the relationships being studied. However, they can only be applied if the variables follow a normal distribution. Since in our case, most variable distributions are not normal (which we will verify using the Shapiro–Wilk test), the entire process of analyzing the test results should be conducted using non-parametric methods. Non-parametric procedures are also more general than parametric ones.

2.4. Results of Fraction Analysis for Survey Responses

As highlighted, the crucial part of the responses lies within the groups of variables P8–P23. We will present what we believe are the most important conclusions resulting from the analysis of the fractions of collected data (though it should be noted that our choice is subjective).

The first issue we want to address is the level of acceptance regarding automation at higher levels. Zoellick et al. (2019), in a study where potential users had prior experience with highly automated vehicles in real-world conditions, found a strong correlation between the perception of safety and acceptance, which was a solid predictor of the intention to use automated vehicles [10]. Therefore, it was assumed that knowledge of assistance systems and their use in vehicles could increase the level of acceptance for automation at higher levels. This issue is related to group P9. As indicated in Table 1, this group included seven subordinate questions with a dichotomous “yes/no” scale. The percentage of “yes” responses is presented in

Table 2. Use of driving assistance systems—question P_9: respondent’s vehicle equipment.

Parking Sensors	Cruise Control	Active Cruise Control	Lane Keep Assistant	Lane Change Assistant	Parking Assistant	No Assistant
In %
45.70	30.80	14.00	15.10	11.60	19.30	36.90

.

The most commonly occurring system is parking sensors, followed by cruise control. Approximately 36.90% of the used cars do not have any assistance systems. More advanced systems are found in an average of 15% of vehicles used by respondents. Considering the low number of vehicles equipped with individual assistance systems, it is challenging to expect a significant impact of this factor on the acceptance of Autonomous Vehicles (AVs). Therefore, the analysis of responses to questions regarding opinions on autonomous vehicles in group P15, which relates to the “level of knowledge”, becomes even more interesting. The percentage distribution of responses is presented in

Table 3. State of knowledge about autonomous cars—question P_15: opinion on autonomous vehicles.

Negative	Rather Negative	Neither Negative, Nor Positive	Rather Positive	Positive
In %
2.6	11.4	30.6	41.4	14.0

.

More than half of the respondents (55.4%) expressed a positive or rather positive opinion, despite the fact that most of them did not use the support systems that characterize the next levels of automation. The group of those with a negative attitude is not too numerous (14%), while 30.6% of respondents had no opinion on the subject.

The level of acceptance may also be due to the level of awareness of the emergence of potential benefits from the implementation and use of autonomous vehicles. Speaking of benefits, they considered, among others, increased driving safety as a result of advanced sensor technologies and AI artificial intelligence.

Table 4. Responses to question P_16: Potential benefits of implementing and using autonomous vehicles (in the group of questions on benefits).

Content	Definitely Yes	Rather Yes	Neither Yes nor No	Rather No	Definitely No
	In %
Increased travel safety	14.9	34.8	29.5	16.6	4.2
Shorter travel time	14.6	30.7	33.2	18.2	3.3
Less traffic congestion in cities	13.2	25.4	30.8	24.1	6.5
New jobs	11.1	25.4	27.1	26.1	10.4

shows responses to selected potential benefits under question P_16 (in the group of questions on benefits) [11].

Nearly half of the respondents (49.7%) see an opportunity to improve travel safety with the implementation of automated and autonomous vehicles, while 45.3% predict shorter travel times. Slightly fewer believe in reduced urban traffic congestion (38.6%) or the creation of new jobs (36.5%). Unfortunately, quite a large group of respondents is not convinced or has no opinion regarding potential benefits, which may also influence the lack of a positive opinion regarding this type of transportation in general. The results indicate a lack of sufficient trust among respondents in these solutions. There is greater agreement among respondents in relation to barriers that limit or slow down the process of transportation automation as shown in

Table 5. Responses to questions from the P_19 topic group: Barriers to the development of transportation based on autonomous vehicles.

Content	Definitely Yes	Rather Yes	Neither Yes nor No	Rather No	Definitely No
	In %
No legal regulations	32.8	42.6	18.6	4.7	1.3
No road infrastructure	35.2	38.7	18.5	6.6	1.0
No service rules	29.3	46.4	18.3	5.2	0.8
No traffic regulations	32.0	41.6	19.0	6.5	0.9

.

The surveyed respondents, in nearly 80% of cases, consider significant barriers to the implementation of Autonomous Vehicles (AVs) to be the lack of legislation regarding the operation of these vehicles, unprepared infrastructure (which serves as a crucial source of information for AV traffic), a lack of service, as well as regulations related to road traffic. These regulatory aspects are particularly important during the transitional phase, i.e., in mixed traffic.

Statistical analysis of the results can help decision makers identify important dependencies and enable them to take appropriate steps, including the dissemination of knowledge about autonomous technologies to potential users. This, in turn, could potentially change their attitudes toward these innovations.

2.5. Results of Variable Dependency Analysis

We analyzed the dependencies between variables using the Spearman rank correlation coefficient, which is a non-parametric measure of the monotonic statistical relationship between random variables. Similar to the Pearson correlation coefficient, it takes values in the range of −1 to 1. We examined the mutual dependencies between variables by testing the null hypothesis of no correlation against the alternative hypothesis of a non-zero correlation, assuming a predefined level of significance α = 0.05. As a result of this procedure, we obtained a correlation matrix and a p-value critical value matrix. Strong correlational dependencies should have a p-value smaller than the predefined α value, indicating that the null hypothesis should be rejected. We will refer to such dependencies as significant. Figure 1 presents a visualization of the absolute values of the coefficients matrix for the analyzed group of variables. To capture more than weak dependencies, we set a threshold value for the coefficient at 0.2 (indicating a more than weak dependency). Coefficients with values lower than 0.2 are not marked in the figure. The remaining coefficients with values above this threshold are shaded in varying degrees of gray proportionate to their significance. It can be observed that there is a significant interdependence between the variable groups from P16 to P23. This suggests that respondents are responding to these questions in a very similar manner, indicating that these questions have identical informational potential.

In the detected groups of variables, one can try to select the group that best characterizes respondents’ opinions. From the analysis conducted in Section 2.4., it was observed that the test results are well described by the group of P15questions: opinion about autonomous vehicles (descriptive statistics of responses to this question are presented in Table 3). In Figure 2, the Spearman rank correlation coefficients for responses to this group of questions are described by column 15 (or row 15—the correlation matrix is symmetric).

In the discussed Figure 2 (and in the

Table 6. Spearman rank coefficients for a group of 15 questions (only more than weak dependence). Row “1”—the name of the group of questions—according to the designation adopted in Table 1. Row “2”—coefficient values indicating more than weak dependence (see the plot in Figure 2 points located outside the lines 0.2 or −0.2).

	1	2	3	4	5	6	7
1	‘P16′	‘P17′	‘P18′	‘P20′	‘P21′	‘P22′	‘P23′
2	−0.5115	−0.4128	0.2825	−0.4505	−0.4169	−0.3327	−0.5231

), one can observe a strong relationship that the responses to this question (information about opinions, e.g., regarding autonomous cars) exhibit with the responses to questions from groups P16 to P23. These groups include potential benefits (P16 questions), acceptance (P17 questions), implementation barriers (P18 questions), personal usage preferences (P19 questions), and travel scenario descriptions (P20 to P23 questions). Since these responses to the above questions are not correlated with other information, such as regarding assistance systems, it was considered that they express a certain stable and independent social view, which may be due to a lack of knowledge on the subject. In our opinion, this view is cautiously accepted. Surprisingly, the opinion regarding the usefulness of autonomous cars is not correlated with the response to question P9 (concerning the equipment of the car with modern assistance systems). Therefore, this result contradicts the view expressed by Zoellick [10]. It seems that the opinion expressed by the respondents is a result of their lack of knowledge about the possibilities of using autonomous cars and is like a preliminary, cautious view on the subject (e.g., “I’m not sure what an autonomous car is, but I’m cautiously in favor”). Responses to the other test questions in the group characterizing the level of knowledge, acceptance, and opinion have a wider range of data and are not correlated. These questions’ responses exhibit greater diversity, leading to the conclusion of an ambiguous assessment of the obtained information.

3. Social (Non)Acceptance in the Area of Automated Vehicle Deployment (Based on Selected Studies)

The presented analysis results indicate a low level of knowledge among respondents. It is widely known that the level of acceptance of any new technology changes as knowledge is acquired. Based on this knowledge, the level of acceptance can either increase or decrease, depending on the nature of the information. The deployment of new technologies is often accompanied by certain benefits, but very often, especially in the initial period, there are also consequences that are not always accepted by society. Predictions regarding potential risks associated with autonomous transportation vary greatly depending on the country, its economic potential, and the level of technology. This is partly due to the lack of experience with autonomous vehicles, as they are still used to a limited extent in transportation systems and are not fully autonomous. Poland’s limited experience, due to its initial phase of activity in this area, requires an analysis and the formulation of forecasts based on the experiences of more advanced countries.

Accurate knowledge and credible presentation of the positive and negative aspects of the implementation of AVs will allow for informed shaping of public opinion. As autonomous transportation develops in Poland, public opinion can gradually change, increasing the level of acceptance for these types of vehicles. The authors have decided to selectively review the results of similar studies in developed countries. Their experiences can be used in the form of “best practices” aimed at minimizing concerns about autonomous and automated vehicles. The analysis of the respondents’ attitudes is often hindered by the fact that fully autonomous vehicles (applies to individual transportation) are not yet available—their implementation is a prospect of several years, so it is impossible to talk about the experience of using them. Conditionally autonomous vehicles, on the other hand, are currently undergoing test studies. It can be assumed that vehicles with conditional automation will be the first to enter into operation. An additional complication is the ambiguous definition of the test object. Questions in the surveys refer to autonomous vehicles, conditionally autonomous vehicles, automated vehicles, or driverless cars. In most of the surveys, these terms are defined, but this does not eliminate the problem associated with the uneven understanding of the phrase “autonomous driving”. This ambiguity can make it difficult for survey respondents to analyze the evaluations obtained and for respondents to form opinions about the obstacles and challenges, as well as the benefits they may gain from using autonomous vehicles. Autonomous vehicles are often seen as a symbol of safe driving because they have the potential to eliminate human error from the driving process. Statistics show that human drivers are responsible for the majority of accidents. This perception of increased safety is one of the potential benefits of autonomous vehicles, and it may contribute to their eventual acceptance once they become more widely available. However, it is essential to address the challenges and concerns associated with autonomous vehicles to ensure their safe and successful integration into transportation systems.

A study conducted at Delft University [13] suggests that achieving social acceptance, particularly in terms of trust in the safety of this new technology, requires the elimination of fatalities caused by malfunctioning AVs, even though AVs currently cause significantly fewer fatalities compared to accidents caused by human error. According to respondents in the study, fatalities resulting from technical malfunctions or software errors are considered four times more significant than fatalities caused by human error. The path to public acceptance of AVs may vary from country to country due to cultural differences related to vehicle usage, which are influenced by national cultural characteristics. Surveys of the German public’s opinion on innovative technologies used in automated and autonomous vehicles were conducted by the FORSA institute on behalf of CosmosDirekt (the main shareholder of the insurance company Generali) [14,15], among others. A market study titled “Expert analysis of the impact of the introduction of automation functions in passenger car fleets on car usability and driving safety” was carried out by the PROGRES institute on behalf of the ADAC company [16]. Both studies addressed similar issues, although the questions were not phrased in identical ways. When comparing the results of the German and Polish surveys, it is evident that 44% of German respondents anticipate an increase in road safety, whereas 49.7% of Polish respondents hold the same expectation. The Polish public appears to be slightly more convinced that autonomy will enhance road safety. This difference is likely influenced by the current lower level of road safety in Poland, which potentially creates a greater opportunity for improvement compared to Germany.

Confidence in the technology was expressed by 27% of German respondents, but only by 12% of Polish respondents. In contrast, a significant 64.8% of Polish respondents believe that it is necessary to maintain control of the vehicle in automated mode at all times. This implies that the majority of respondents in both Poland and Germany do not have full trust in autonomous cars or new technologies that enable autonomous driving. However, the level of trust appears to be lower in Poland. In an international study covering Austria, Germany, South Korea, and the USA, almost half of the respondents identified trust as the primary barrier to the acceptance of Autonomous Vehicles (AVs) [17]. Confidence can be influenced by two key factors: prior experience and system transparency. A simulated experiment involving partially automated vehicles demonstrated that positive experiences increased trust in the vehicle, while negative experiences (such as accidents) decreased trust [18]. System transparency, which refers to the extent to which users can predict and comprehend vehicle performance, was found to positively impact trust in AVs in a study that applied the Technology Acceptance Model to AVs [19].

A study involving 2000 respondents from the U.S. and Germany revealed that individuals who have at least one Advanced Driver Assistance System (ADAS) feature in their vehicles are much more likely to consider using both private autonomous vehicles and on-demand autonomous vehicles in the future, with percentages of 68% and 65%, respectively. Conversely, those who do not own cars or lack advanced assistance systems in their vehicles are much less inclined to intend to use both private autonomous vehicles and on-demand vehicles in the future, with percentages of 42% and 22%, respectively. Unfortunately, such a relationship was not observed in the Polish study due to the limited number of vehicles equipped with advanced technology.

4. Discussion

Acceptance is a dynamic concept, heavily influenced by people’s attitudes, expectations, behaviours, environment, established values, norms, and knowledge levels, among other factors. The level of acceptance can evolve over time and is greatly dependent on various attributes, types, subject matter, and the specific context in which this attitude is being examined. Currently, on a global scale, the applied technologies in autonomous vehicles rank among the most researched and sought-after topics, not only among vehicle manufacturers but also among prospective users.

Vehicles are progressively incorporating automation through the introduction of smart features such as adaptive cruise control (ACC), lane-keeping assist (LKA), and driver monitoring for attentiveness and health issues. However, fully autonomous vehicles are not yet available as consumer products in many countries, and there remains substantial scepticism among the public regarding autonomous vehicles. Therefore, building public trust in these modern technologies represents a significant challenge. It is trust that ultimately determines the success of deployment and the willingness of prospective users to embrace and utilize them [20,21]. Distrust in machines does not solely arise from concerns about the reliability of sensor technologies but is also rooted in various factors, including a fear of losing control over one’s own destiny. This loss of control represents one of the fundamental philosophical issues in the context of artificial intelligence development. Consequently, it is crucial for sociologists and psychologists to conduct in-depth research into this fear and formulate strategies for educating the public.

As a result, vehicle manufacturers and governments are actively seeking answers to the following question: What steps should be taken to enhance the confidence of users in future technology? This inquiry is at the forefront of efforts to ensure the successful adoption and acceptance of advanced technologies, including those related to autonomous vehicles and artificial intelligence.

The findings from Yang’s research [22] on the acceptance of autonomous vehicles revealed a pattern of trust that evolves and stabilizes over time. Trust tends to grow as users have the opportunity to directly engage with new solutions, enabling them to assess these technologies based on their current and past experiences with previously available driver assistance systems [20]. Positive evaluations from users can lead to increased trust in the technology and greater acceptance of it [23]. Additionally, users expressed the belief that future generations may be more at ease with autonomous vehicles and their features, as automation becomes a part of their lives from an early age. This suggests that familiarity and positive experiences play a significant role in building trust and acceptance of autonomous vehicle technologies.

It is noteworthy that trust and receptivity to new technologies tend to vary inversely with the age of users. Younger generations exhibit significantly greater confidence in fully autonomous vehicles compared to their older counterparts. For instance, 56% of vehicle owners belonging to “Generation Y”, often referred to as millennials, express trust in self-driving technology, and a similar sentiment is shared by 55% of “Generation Z” individuals (those born in the mid-1990s or later) [24].

One common concern shared across all generations pertains to the security of the technology, especially issues related to privacy, potential hacking, takeovers, or system failures, whether of the vehicle itself or the underlying technology [25]. This research suggests that acceptance of autonomous vehicles will naturally increase over time as new generations replace older ones, which is an optimistic outlook for the future of this technology.

In contrast, a Deloitte report indicates that potential users are inclined to trust autonomous vehicles when there is a well-established safety track record. This trend is evident in Belgium, where an increasing number of consumers are looking to reputable car manufacturers for assurance regarding the safety of autonomous technologies. For instance, 58% of Belgian consumers state that they would be more inclined to use an autonomous vehicle if it is produced by a brand they trust [26]. It suggests that trust in established brands plays a pivotal role in fostering confidence in autonomous vehicle technology. Experience also plays a major role in this case.

The acceptance, trust, satisfaction, and perceived usefulness of autonomous vehicles (AVs) demonstrated a significant increase after participants had experienced a driving simulator, as compared to their initial attitudes [23].

Lustgarten’s research highlights that in a questionnaire, more than half of the participants prioritized the ‘highest possible level of safety’ when it comes to automated cars [27]. Consequently, manufacturers of autonomous vehicles should give special attention to ensuring the safety of autonomous cars and should work to demonstrate to the public that driving an autonomous vehicle is not a risky endeavor [28]. This emphasis on safety is crucial in building trust and acceptance among potential users.

In Poland, efforts have been made to address recommendations regarding the introduction of autonomous vehicles on the roads. These recommendations were developed as part of the mentioned AV-PL-Road project and documented in the Roadmap, with the article’s authors being among the contributors. The recommendations covered a wide range of aspects, not only concerning increasing the level of public acceptance. Specifically, the following were included:

Introducing regulations enabling safe testing of autonomous vehicles on roads (amendment of the Road Traffic Law).

Implementing a requirement for testing autonomous vehicles on tracks before allowing them on the roads.

Creating a database of scenarios for simulation testing.

Training highly specialized personnel.

Educating drivers and driver candidates and enhancing societal acceptance through the promotion of automated transport.

Regarding actions to increase the level of trust in Polish society, as indicated by research, it is crucial to promote accurate information about autonomous vehicles and promote it appropriately. There are various tools and forms of promotion that can be utilized, such as:

• Advertisements.
• Extensive social campaigns.
• Social media.
• Partnerships.
• Traditional media.
• Events.
• Real-time marketing.
• Education of specialists.
• Public education.

The choice of tools may be guided by various criteria, with the key objective being to reach as many potential users as possible. Therefore, the key to success lies in disseminating knowledge among future users of autonomous vehicles and specialized personnel.

It is worth noting that some of these actions have already been taken, such as organizing three editions of the AV-Poland conference in collaboration with the Warsaw University of Technology Faculty of Transport and the Institute of Automotive Transport in 2017, 2021, and 2022, addressing the market of autonomous vehicles in Poland.

5. Conclusions

The survey aimed to assess public perception of new technologies in autonomous vehicles. The analysis of the results indicated a low level of knowledge among respondents. As known, acceptance of new technologies changes with the acquisition of knowledge. Therefore, a priority is to undertake initiatives that build trust in autonomous vehicles, thereby contributing to sustainable development, improving transportation efficiency, reducing negative environmental impact, and increasing accessibility of transportation services for various social groups.

Future directions of work may include the following actions. First, there is a need to prepare an educational campaign to inform residents about autonomous technologies. The results presented in the article highlight the necessity of clearly explaining to society issues related to safety, efficiency, and potential benefits of autonomous vehicles, such as reducing road accidents, improving public transportation accessibility, and reducing air pollution. Collaboration with local authorities and communities is crucial in this process to better understand concerns and adapt technology to local needs. Consultations with residents should also be conducted to incorporate their opinions and feedback.

From the perspective of residents or passengers, the key issue is the safety of autonomous vehicles. Providing information about the principles of designing new technology, its maturity level, safety testing, and research results will positively impact the level of trust in autonomous vehicle technology among the public. Efforts should be made to demonstrate ongoing technological successes by organizing test rides of autonomous vehicles in real or simulated conditions. Direct contact with the technology can help residents dispel concerns and uncertainties about its actual capabilities. Transparency in the use of solutions also positively influences the perception of the entire technology. Therefore, the use of open-source software solutions and tools is recommended to gain support from the global IT community and should form the foundation for a future unified hardware-software platform for autonomous vehicles.

The challenge of establishing full confidence in the technologies employed in autonomous vehicles is multifaceted. Nevertheless, striving to achieve the highest possible level of public acceptance is a top priority.

In summary, providing clear and understandable information to the general public about the operation, safety, and benefits of autonomous technology, along with educational initiatives such as training programs, presentations, and educational videos, can help alleviate concerns about the safety of autonomous driving and increase acceptance of this new technology. Such strategies are outlined in the Autonomous Vehicles Road Map for Poland, indicating a concerted effort to promote acceptance and understanding of autonomous vehicles.

In summary, these initiatives are aimed at overcoming obstacles associated with the lack of social acceptance of autonomous vehicle technology by increasing knowledge, addressing concerns, and promoting the benefits of their use. These actions are crucial for building full trust in autonomous vehicles and achieving widespread social acceptance.