The insights from research into car usage and ownership, together with the indications from the first exploratory study on autonomous driving, have formed the conceptual basis for our follow-up empirical work. The objective of these ongoing studies is to survey, in a use-oriented fashion, the specific attributes that road users currently ascribe to autonomous cars and, in the “translation” of such attributes, to decipher instrumental, affective and symbolic motives. The questions guiding the research are thus: What characteristics and assessments do road users associate with autonomous driving and autonomous vehicles? What various motives are decisive in this? The context of what attributes are currently ascribed to autonomous driving is also to be analyzed and, where possible, linked to how people live their daily lives and get around. In this way, it will then be possible to determine which attitudes and judgments may be expected to have an effect on acceptance of the technology.
To this end, a multiple-method procedure was selected. To explore potential variations in people’s perceptions and judgments, respondents were confronted with specific use cases of autonomous driving in a quantitative online survey. This survey also contained a free-text questionnaire—participants were thus able to ascribe autonomous driving with attributes in a completely individual way. A more in-depth analysis encompassing the context in which autonomous driving is presently perceived and viewed allowed for exploratory qualitative procedures to be pursued in parallel as part of group discussions. The focus here lay particularly on ambivalence in how technology is viewed (see Chap. 29), particularly in advance of potential implementation—with both fear and high expectations. Exploratory qualitative procedures appear to be especially apt for this still relatively young field of research. Until now, there has been barely any insight available into the attitudes and motives surrounding autonomous driving on the user side—the open character of such studies helps to decipher these motivations.
Perception and Assessment of Autonomous Driving in Relation to Specific Use Cases
It may also be expected that each of the four use cases developed in the course of this project (Chap. 2) will be perceived and assessed differently in view of their acceptance-relevant aspects. This is because they not only differ in technical terms, but also in the specific implications, usage areas, and attributes they are associated with (for more on this see Chaps. 6, 11, 12, and 32). In a quasi-representative survey, we therefore recorded differentiated attitudes to autonomous driving with recourse to the use cases. A thousand people were surveyed, and represented the total German population regarding gender, age, income, and level of education. An in-depth discussion of the survey—which was developed and carried out in cooperation with several authors of the present volume—and sampling can be found in Chap. 6.
A total of 57 % of those surveyed declared they were generally interested in the topic of autonomous driving. However, 44 % stated they had no knowledge of the subject, and a mere 4 % termed themselves as being well informed or having specialist knowledge at all, let alone being an expert. 78 % obtain their information on the topic primarily from the mass media, 64 % go straight to experts, 56 % discuss it with friends or co-workers, and 40 % share and compare notes on social media.
Following a general section with questions on sociodemographics, mobility behavior, travel requirements, etc. the respondents were each randomly allocated one of four scenarios based on the use cases. Each scenario was introduced with a short description:
Scenario for Interstate Pilot Using Driver for Extended Availability: On freeways or similar roadways, the driver can hand over control to the vehicle. The driver does not have to pay attention to other traffic or the driving task in this time, and can pursue other activities.
Scenario for Autonomous Valet Parking: After all passengers have got out of the vehicle, it can drive itself to a pre-determined parking space and also from there to a pick-up address.
Scenario for Full Automation Using Driver for Extended Availability: Whenever and wherever desired or required, the driver can hand over control to the vehicle. The driver does not have to pay attention to other traffic or the driving task in this time, and can pursue other activities.
Scenario for Vehicle on Demand: A vehicle on demand drives its passengers without the presence of any human driver. People themselves can no longer drive in such vehicles – the vehicle’s interior will thus also have neither steering wheel nor pedals.
A follow-up question to each use case as it was introduced enquired whether the respondents would essentially be prepared to replace their till-now preferred transport mode with an autonomous vehicle. This question has had already been put to them in the same form in the first, general part of the survey, but there it was relatively unspecifically termed “an autonomous vehicle” with no further explanation. Altogether the respondents had relatively little desire to replace their own vehicle (or “favorite mode of transport”) with an autonomous vehicle—whether precisely specified or not. Only between 11 and 15 % agreed to this statement in large or full measure (see Fig. 31.1). However, 27 % said they could hardly imagine, or could not imagine at all, replacing their preferred mode with a (non-specified) autonomous vehicle. When, as in the survey, autonomous driving is proposed in relation to a specific use case, this degree of refusal increases sharply, to between 44 and 54 %. This means that refusal becomes stronger with more precise scenario descriptions. The lowest acceptance, incidentally, is for Vehicle on Demand—54 % would not wish to replace their favored mode with it and only 11 % could envisage it at all.
To explore what the respondents currently associate with autonomous vehicles, they were asked to declare in their own words, in up to fifteen free-text boxes, what they understand by the term “autonomous vehicle.” The short descriptions (see above) were also a basis here. The following analysis refers solely to the answers of those respondents who had been allocated the “Full Automation Using Driver for Extended Availability” (referred to as “Fully automated vehicle” in the following) and “Vehicle on Demand” use cases.
The answers of the 250 respondents were summarized and categorized by hand, then allotted specific connotations (see below). For “Fully automated vehicle” there were a total of 3750 entries; of these, 2587 (69 %) were invalid for various reasons because it was not possible to make out a reference to the question, for example. For Vehicle on Demand, there were also 3750 entries overall, of which 2512 (67 %) were unusable. Figure 31.2 shows the distribution of statements with various connotations: positive, ambivalent, negative or without connotation—the invalid entries have already been taken out at this point, the percentages refer to the remaining mentions.
While the majority of descriptions of Fully automated vehicle are positive, the same can only be said of 38 % of definitions applied to Vehicle on Demand. For these two use cases, 36 and 40 % of the statements respectively had negative connotations. A small portion of the perceptions (5 and 4 % respectively) were ambivalent, i.e. they could not clearly be connoted as positive or negative.
Overall, the qualities that the respondents attributed, independently of one another, to each of the two scenarios turned out to be relatively similar. Many answer categories are both equivalent in meaning and similar in their percentage distribution. We shall turn next to some of the notable differences. We shall also examine the attributes given, where possible, for what instrumental, emotional, and symbolic characteristics are ascribed to autonomous vehicles.
Full Automation Using Driver for Extended Availability
In the “positive assessments” segment, 17 % of statements were in the category of “comfortable,” followed by “good” (13 %), “safe” (11 %), “relaxing” (10 %), and “modern” (10 %)—these proportions are given in Table 31.1. Among the “ambivalent assessments,” 8 % of the answers came in the “luxury” category, and 15 % of all “negative assessments” were because they viewed the scenarios as “expensive.” In this negative segment, only the answer category of “not for me” was more highly represented, at 16 %. “Luxury” was not once mentioned in connection with Vehicle on Demand, however, and, at 7 % of all statements, “expensive” only came in at seventh place.
In the positive assessments segment, Fully automated vehicle was mainly associated with functional or instrumental aspects—vehicles such as these are described as “comfortable,” “safe,” “practical,” “efficient,” “for the mobility-impaired,” “helpful,” “environmentally friendly,” and “flexible.” Vehicle on Demand was also deemed “useful,” although no-one deemed it “helpful.” Regarding Fully automated vehicle, only the positive attributes of “relaxing,” “brilliant,” and “exciting” could be termed as emotional or affective. The same applies for Vehicle on Demand, though the relatively weaker connotations of “great” and “good” categories were expressed instead of “brilliant.” As for the negative assessments, on the other hand, the distribution of functionally versus emotionally connoted aspects is reversed. “Weird,” “boring,” “dangerous,” and “terrible” are clearly affective categories (and are joined in the case of Vehicle on Demand by “scary”), while “expensive” and “insufficiently developed” denote more functional aspects. Answer categories that may be assumed to have a symbolic connotation are scarcely to be found among the statements. Statements in the “modern,” “interesting” or “luxury” categories most likely indicate that such autonomous vehicles are perceived and assessed in terms of aspects concerning the status manifested in the vehicles themselves.
Vehicle on Demand
The summarized and categorized attributes given for the Vehicle on Demand can be seen in percentage in Table 31.2. The top answer categories in the area of “positive assessments” hardly differ from those for Fully automated vehicle. Only in first place do we see a completely new category of “useful,” with 15 % of all statements made in this segment. After this, the respondents go on to describe the vehicle as “comfortable” (14 %), “relaxing” (13 %), “modern” (12 %), and “safe” (10 %). Eighteen percent of statements made about Vehicle on Demand have no judgment attached to them. Half of these come under the “no idea” category—only 20 % of respondents, in comparison, declared they had “no idea” about Fully automated vehicle. Only 2 % of statements in the “no valuation” column were in the category of “similar to other transport modes,” whereas this applied to 12 % of those made about Fully automated vehicle. This category included entries such as “Fully automated vehicle is ‘like the railroad’” or “Vehicle on Demand is ‘a taxi.’”
A quarter of all negative statements on Vehicle on Demand fall under the highly emotionally cast categories of “scary” (10 %), “dangerous” (7 %), “weird” (6 %), and “terrible” (2 %). In contrast, not a single respondent found Fully automated vehicle to be “scary,” and only 15 % termed it “weird” (11 %), “dangerous” (11 %), or “terrible” (1 %).
Overall the survey clearly demonstrates that Vehicle on Demand is the subject of the greatest number of negative and fewest positive assessments. Of the 250 respondents allocated to this use case, 54 % could not imagine replacing their currently preferred mode of transport with a Vehicle on Demand. In direct comparison with Full Automation Using Driver for Extended Availability, Vehicle on Demand is described with rather negative statements, with a quarter of all descriptions even viewing it as scary, dangerous, weird, or terrible. Clearly, this means that fewer respondents could imagine using a Vehicle on Demand than one they can still drive themselves, as is shown by the number of statements professing “no idea.”
The answer categories of “expensive” and “luxury,” which are either especially or exclusively attributed to Full Automation, indicate that such vehicles are still clearly linked with individual private ownership. Vehicle on Demand, on the other hand, is compared with other transport modes little or not at all.
Classifying the statements on autonomous driving in terms of their instrumental, affective or symbolic aspects revealed that instrumental attributes predominate among positively connoted assessments. Strongly emotional statements, on the other hand, form the majority of negative accounts. Descriptions foregrounding the status character of autonomous vehicles, in contrast, were scarcely to be found. In the qualitative survey described in the following section, we shall take a closer look at the perceived negative aspects of autonomous driving, and ask which sociodemographic contexts they are embedded in. Such hostile attributes, in particular, are evidently strongly aligned with subject-related and affective issues as well as the context of car use and ownership, as the exploratory study has already shown (Chap. 29).
Autonomous Driving in the Future: “Do We Really Want to Live Like That?”
The following findings are based on three group discussions in Berlin whose composition varied each time. All participants had a high level of education; some were in academia, studying or working at a university or other research institutes. The discussions involved five, six, and seven participants respectively. All those involved were living in Berlin, seven of them were women. The gap between oldest and youngest was greatest in the first discussion round, with the youngest 20 years of age, the oldest 50. All participants regularly used a car, though not all of them owned one. At the end of each discussion, data on sociodemographics, transport behavior, and car use and ownership were collected. All those taking part had already heard of autonomous driving before the discussion sessions.
At the beginning of the session, the participants were given an illustrated narrative scenario of autonomous driving in the form of an A4 flyer. There were two different scenarios—one on “Full Automation Using Driver for Extended Availability,” the other on “Vehicle on Demand.” Each group was introduced to one of the scenarios, which they then considered within group-based introspection. The aim of this introspection is to make explicit the processes of “inward observation”, that is to say, to consciously direct concentration and attention towards interior processes (: 493, translation by the authors). Under the guidance of a research person, a group’s participants address the research object in question and document their own inner processes and experiences independently of one another. Finally, they share what they have found via self-observation with the group without commenting or passing judgment on each other’s accounts. In a second stage, the group members—keen explore the topic further from hearing others’ experiences—add to their reports. Together with the scenario, the participants received the following instructions:
You have just received a short description of a scenario involving the cars of the future – a short story on what the driving of tomorrow might look like. Please read the story carefully. Try to imagine yourself in what you read and think about your feelings, fantasies, and sensations – be open for anything and everything that crops up as you think about the topic!
Please make notes on where your thoughts lead you.
The participants then each read out their notes in turn, as a so-called “introspection report.” The reports are later transcribed and evaluated by means of qualitative heuristics. Unlike everyday procedures of discovery, which are frequently made unconsciously, qualitative heuristics is “rule-directed and supplied with a methodology,” and takes the form of systematized and intersubjectively traceable search-and-find procedures (: 226, translation by the authors). The method is based on four rules:
Openness of the research person and subjects
Openness of the research object
Maximum structural variation of perspectives
Analysis of commonalities.
Furthermore, qualitative heuristics uses the so-called dialogic principle: A question is asked of an object (in the study, the transcribed introspection reports) which gives “answers.” New questions are then asked from another perspective, or a different angle, and so the process continues. The research object and research person are thus in close dialogic contact, which also serves to soften the strict division between (research) subject and (research) object.
After producing introspection reports, the groups started open discussions, which were aligned on the implicit behavioral patterns (see ) in car usage and ownership. In now giving our results, however, we shall focus on what the participants conveyed in their introspection reports.
The scenarios allocated to the group discussion participants stimulated great interest in the topic, but also critical questioning. In what follows, we shall describe some examples of topics that had already proved significant in the earlier, exploratory survey. The introspection reports show an ambivalent attitude to autonomous driving that is comparable to the results of the study given in Chap. 29.
In addition, however, it became clearer which specific fears and worries relate to autonomous driving, and what social context it is seen in. We shall take a closer look at the range of hostility towards autonomous driving below—the results stem from both groups who were given the “Full Automation Using Driver for Extended Availability” scenario.
Skepticism About a Future with Autonomous Vehicles
For the sake of completeness, it should be mentioned at this point that, in the scenario stories provided to the participants, the protagonist “Yvonne” used the time made free to her from not having to drive herself to pursue work activities (e.g. email correspondence), among other things. Although other activities were also covered alongside this, the story did tend to focus more on typical organizational activities (taking the kids to school, doing the shopping, etc.) than on leisure and relaxation pursuits (looking out the window, watching movies, sleeping/relaxing, etc.). It is therefore possible that skeptical and hostile attitudes on the part of (especially younger) group members towards autonomous driving could be down to the fact that life in the future was so clearly depicted as being full of structured and optimized everyday tasks.
On the other hand, the qualities and attributes ascribed to autonomous driving by the participants can clearly be classified in terms of the earlier surveys, with the same ascriptions cropping up again and again in both the exploratory study and the quantitative questionnaire.
The “accelerated service society”
That it will in future be possible to spend time on other activities in an autonomous vehicle was largely viewed negatively in the discussions. In their orally read-out introspection reports, the participants expressed concern about no longer needing to concentrate on the driving task in future. This, they thought, could lead to private, leisure, and work activities becoming too closely mixed up with one another. Ultimately, this line of thought runs, technology may further a trend that many people today already identify as alarming: a society ever-more oriented towards performance and efficiency:
The names of the participants have been changed to protect their anonymity, and the quotes have been translated.
“And private life and working life are mixing more and more and you become a total workaholic.”
“This strain of having to do more and more things in the same place at the same time is increasing.”
The freedom and the opportunity to occupy oneself during the car journey in other ways may result in pressure to put this time in the service of efficiency. Having to concentrate while driving conventionally, on the other hand, was painted in a positive light:
“That’s actually a nice thing about driving, that you have to concentrate on it in the moment and you also do something with your hands, and you precisely aren’t already checking emails from work. That starts when you’re sat at your desk.”
“This dependency on technology”
One further consequence linked to autonomous driving is a future dependency on technology, which may also entail an accompanying large degree of loss of control, which is viewed negatively. Technology dependence and loss of control are also seen as problematic because there is skepticism as to the reliability of a technology over whose decisions one will no longer have any influence:
“In certain situations you can simply decide spontaneously, and you might actually have a much better feeling for it than the car.”
“And obviously it won’t be possible to intervene immediately either if I first have to hear 5, 4, 3, 2, 1…beep.”
Evidently, behind such worries lies a fundamental skepticism regarding technology, all the more so when it greatly impinges on individual safety, as in the case of autonomous driving:
“Not even my kettle gets my blind trust, why should I then blindly trust my car with my life? I somehow find that incredibly disconcerting.”
Life is being “de-funned” and “you’ll get idle”
The abolition of the driving task in autonomous vehicles is viewed highly critically. The main idea here is that a vehicle in which one does not drive oneself would curtail fun, spontaneity, individuality, flexibility, and control (it is interesting to note here that the participants had actually been given the Full Automation with Driver for Extended Availability story, where—as is described in the scenario—the driver can certainly still drive themselves whenever they like):
“What is actually then the difference to public transportation? Because what I actually always appreciate about a car is that it is in my own hands—that I can judge for myself. And if I’m running a little late, then I can step on the gas a little.”
The “de-funning” (Timo) comment refers on the one hand very specifically to losing the fun of driving, but also to the lifestyle that could have the fun removed from it by autonomous driving:
“You’ll stop needing to move so much because you’ll be able to get picked up by the car everywhere; you’ll get lazy, you’ll get idle, you’ll only ever take the car, because it doesn’t matter if you’re feeling ill or how you are – you can take the car for every trip you make.”
Social isolation: “Nobody needs anybody anymore”.
The consequence of such “de-funning” and idleness is, in the view of the group members, that humans in the end will be replaced. “Nobody needs anybody anymore” (Inga), there would then be a machine for everything and, in this way, even thinking would be done for you:
“The car drives, you no longer have to drive yourself. Food is delivered in one way or another, you cower a touch autistically in your apartment and grow dull, you don’t have to think anymore, you google or subject yourself to cat videos. So you basically become completely stupid.”
System weaknesses: How truly “autonomous” are autonomous cars?
Overall, trust in the safety of autonomous vehicles appears to still be on the low side. This leads the participants into wide-ranging speculation: “How predictable are these cars, then?” and, “What about when the system breaks down and there is no internet access?” (Nico), “Can they be driven remotely?” (Thorsten), “What happens when this system gets hacked?” and, “Can these autonomous cars really be autonomous?” (Bettina), “Just where, then, is the evidence that the whole thing is going to be safe?” (Herta).
Behind such questions and statements, there is great insecurity regarding the still-unknown technology, and the possibility that dangers may accompany it. At the same time, the participants currently appear to be completely unclear as to who is behind the development of this technology and who is responsible for the system’s safety. Who would be liable in case of an accident? Who would accept the damages? And most important: Who would accept responsibility in the ethical sense?
“Social and economic consequences”
The examination of autonomous driving also turns up the worry that the technology will lead to job losses in various sectors (the automotive industry, taxi and delivery services, etc.). This is connected to the idea that autonomous vehicles will result in less variety across the spectrum of automobiles: “If you only have this large one” (Bettina), autonomous driving would in the end be accompanied by “greater monopolization” (Eddie).
The issues raised in the group discussions concerning autonomous driving are embedded in a sociotechnological context that simultaneously uncovers negative and problem-centered ideas of society in future. Overall, the participants tend to take a hostile position vis-à-vis autonomous driving. Although positive aspects are also perceived, these are neither embedded in a specific usage context nor associated with positive ideas of future society—at least not to the same extent as the negative ones.
Fears expressed in connection to a future with autonomous driving concern social isolation, social and economic consequences, over-reliance on technology, increasing idleness, and the pressure to keep up in a primarily performance-oriented society.