Towards Theorizing Co-Driving Pleasure

Research question: from driving to co-driving pleasure

Driving pleasure has a long history in automotive design, often associated with a person’s direct control of the car. Driving automation changes that practice: a driver may delegate actions, monitor what the system is doing, and take back control as the situation changes.

My doctoral research investigated co-driving as this collaboration between a human driver and driving automation. Rather than beginning with a new interface, I began with an experiential question:

How does pleasure emerge in everyday co-driving, and how can understanding it inform design?

The systematic literature review identified a conceptual gap: discussions of pleasure around automated vehicles often carried forward assumptions from manual driving. Reducing the discomfort of lost control does not, by itself, explain what people find pleasurable when control is shared. I wanted to understand the activities, situations, and relations through which that experience takes shape.

Define the scope before studying the experience

The inquiry focused on advanced driver-assistance systems (ADAS), rather than fully driverless travel. From the driver’s perspective, the research organized features into three groups:

Control symbol used to represent automated vehicle control — Vehicle control: features that act on vehicle motion or controls.

Warning symbol used to represent emergency alerts — Emergency warning: features that alert the driver to a potential hazard.

Information symbol used to represent enhanced environmental information — Information enhancement: features that help drivers perceive their surroundings.

This distinction mattered because “automation” was not one uniform experience. Delegating motion, receiving a warning, and consulting information can involve different expectations and different ways of relating to the car.

Research design: connect experience, interpretation, and theory

The study combined trip experience sampling, two rounds of phenomenological interviews, qualitative interpretation, and discussions with automotive design practitioners.

Research design connecting first-round experience sampling and interviews, three themes, theoretical lenses, model building, second-round triangulation, and practitioner feedback — The research sequence: experiential inquiry → themes → theoretical interpretation → model building, with a second interview round and practitioner discussion.

The first round established an experiential foundation: 12 participants submitted 115 trip logs and took part in interviews. Analysis developed three themes. Technological Mediation Theory and pleasure theories in design then supported conceptualization of the Human–Vehicle Relation (HVR) framework and Co-Driving Pleasure Model.

A second round involved 10 participants and provided a further source of accounts for examining the themes and model. Finally, nine automotive design practitioners discussed the research and its implications for design practice. These stages served different purposes: gathering experience, interpreting it, developing concepts, and examining their relevance to practice.

Fieldwork: ground interviews in particular journeys

Recruit people who already co-drive

Recruitment combined convenience and criterion sampling through social media and field contact, including a Tesla Supercharging Station in Turin. Participants needed to drive regularly and use driving automation in everyday trips.

Tesla Supercharging Station in Turin used as a field recruitment location — One field recruitment setting in Turin, Italy.

The first interview round primarily involved Tesla drivers. The second broadened the vehicle range, including NIO and BMW, to bring further accounts into the inquiry. This was a qualitative study of lived experience, not a representative survey of vehicle owners.

Capture experience close to the trip

Participants were asked to complete 10 logs, recording their subjective experience shortly after each journey. They could use either a paper booklet or a QR code linking to Microsoft Forms; both formats contained the same questions.

Paper trip-log booklets alongside QR-code cards linking to the digital form — Two formats for the same experience-sampling instrument.

Original trip-log instrument recording trip purpose, context, experience, automation use, and evaluation — The original instrument made a specific trip available for later discussion. The image can be enlarged to read the questions.

The 12 first-round participants submitted 115 logs. I organized and categorized the records in Excel to identify recurring driving contexts and feature-use patterns, then used those patterns to tailor the subsequent interview. The logs were a way into concrete episodes, not a substitute for exploring what those episodes meant to the driver.

Move from a record to an experiential account

The semi-structured phenomenological interviews lasted 40–90 minutes. They explored the participant’s driving patterns, their understanding of pleasure in each context, and the role of particular assistance features.

Interview preparation sequence from recruitment and trip sampling through driving-context sorting to tailored semi-structured interviews and transcripts — Trip records informed the questions, while interviews gave participants room to confirm, challenge, and explain the contextual patterns.

Prompts moved between the wider account and particular episodes:

What kind of driver do you consider yourself, and does the proposed grouping of your driving contexts make sense?
What do you normally do in this situation, and what does driving pleasure mean here?
Why do you use this assistance feature in this context? What experience does it give you?
What criteria do you use when judging how the feature performs?

The intention was to understand the experience in its circumstances, rather than collect a general opinion about automation.

Analysis: move between the whole account and its parts

I followed a whole–part–whole interpretive process. The transcribed interview provided an initial whole. Driving-context-related and ADAS-related passages formed a preliminary structure for examining particular experiences. I then returned to the accounts to develop a new whole: themes of co-driving practice and pleasure.

Whole–part–whole interpretation moving from transcribed interviews through contextual and ADAS-related parts to themes of co-driving pleasure — Interpretation was iterative: the meaning of an episode was considered alongside the participant's wider account.

The sampling records and interviews also offered different views of feature use. A feature could appear in a trip log, in an interview, or in both. Putting these sources alongside one another helped organize the material without treating the frequency of mention as the meaning of the experience.

Original matrix comparing ADAS features across driving-context categories, using participant codes and distinctions between trip-log and interview evidence — An analytical working view of accumulated experience patterns. Codes refer to participants in the research; the matrix is not a population-level measure.

This process produced three connected themes: how co-driving is practiced, what contextual factors influence pleasure, and the forms that pleasure can take.

Findings: three themes of co-driving experience

1. Co-driving develops through trust, knowledge, and habits

Drivers did not simply adopt all available automation. Their accounts revealed selective use and different judgments about when delegation was appropriate.

Trust was a threshold. Giving a system control of vehicle motion could raise different concerns from receiving information or warnings. Mutual knowledge mattered: drivers’ understanding of a feature and judgments about its ability to handle a situation shaped its use.

Habits could develop or conflict. Automation could become part of a new routine, while an established driving style could also lead someone to resist a feature. Evaluation therefore concerned both practical usefulness and how the system exercised its autonomy.

These findings directed attention away from the presence of a feature alone and toward the relation that develops through its use.

2. Five contextual factors shape the possibility of pleasure

Five influencing factors of co-driving pleasure: spatio-temporal conditions, driver status, affective factors, instrumental factors, and socio-cultural factors — Pleasure is situated: the same feature is encountered within a changing configuration of circumstances.

Spatio-temporal conditions: the road, other road users, weather, and time-related circumstances of the trip.
Driver’s status: mental and emotional states before and during the journey.
Affective factors: feelings toward driving and attachment to the car.
Instrumental factors: the trip’s aim and the driver’s knowledge and skills.
Socio-cultural factors: driving style and habits, the image of the car or brand, and trust in ADAS.

Together, these factors help explain why a feature can contribute to pleasure in one situation but not another. They are contextual dimensions for inquiry, not a formula that predicts an emotional outcome.

3. Pleasure includes both a good mood and pleasure with the car

The interviews extended pleasure beyond the familiar image of exciting manual driving. Two main forms encompassed five subtypes:

Research illustration of good mood during co-driving, including anticipation and social experience — Type I — Good mood as co-driving pleasure.

Research illustration of pleasure with the car through reduced effort and embodied movement — Type II — Pleasure with the car.

Good mood included the continuation of a mood that existed before the trip, anticipation of the destination, and joy arising from social interaction. Here, pleasure was not necessarily produced by the automation itself; the journey could sustain or accommodate it.

Pleasure with the car included relaxation through reduced mental or physical effort, and the embodied experience of moving with a responsive vehicle. These accounts connected pleasure more directly to what the car enabled the person to do and feel.

The distinction raised a conceptual question: how do situated activities, fulfilled needs, and relations with the vehicle come together in a pleasurable experience?

Conceptualization: a relational account of the vehicle

Three experiential themes interpreted through theoretical lenses to develop the Human–Vehicle Relation framework and Co-Driving Pleasure Model — The themes were a foundation for theoretical interpretation, rather than the finished model.

Two lenses supported this step. Technological Mediation Theory helped examine how a vehicle shapes action and perception. Pleasure theories in design supported an account of pleasure in relation to the fulfillment of needs during activity.

Human–Vehicle Relation: technology participates in experience

I used Human–Vehicle Relation (HVR) to describe the dynamic, evolving relation between a driver and an automated vehicle. The vehicle is not only an object with useful attributes; it mediates how the driver acts and experiences a situation.

Human–technology–world diagram showing mediation of perception and experience, and of actions and practices — Technological mediation: technology can amplify or reduce perception and invite or inhibit action. The original diagram cites Verbeek (2005), adapted from Hauser et al. (2018).

This changed the unit of inquiry. Instead of asking only whether a feature is usable, I could ask what kind of relation it makes possible, how that relation changes with context, and how it participates in a wider experience.

Compose the wider relation from feature-level relations

Drawing on Rosenberger’s organization of experience and an analogy with service-design touchpoints, I introduced Human–Feature Relations to articulate how individual features contribute to a wider HVR.

HVR composition diagram showing multiple feature relations through field composition, sedimentation, transparency, and forefronting — The composition diagram articulates how several feature-level relations can coexist in one experience.

The diagram distinguishes four aspects: field composition, how technologies shape the field of awareness; sedimentation, how use becomes habitual; transparency, how a technology can recede from explicit perception; and forefronting, how some relations become more salient among others.

This made the broader relationship discussable without reducing it to a list of isolated functions. It also provided a route toward the later Mediation Composition Diagram project.

Distinguish design intention from the relation experienced in use

The Inscribing and Rendering Model addresses the designer’s role. Designers embed intended relations in a product, but drivers encounter and enact that product within their own situated practices.

Inscribing and Rendering Model linking a designer's intended human–vehicle relations, product meaning, and relations enacted by the driver — Intended relations are inscribed through design; experienced relations are rendered through use. They need not coincide.

The distinction matters for design research: an intended form of support is not evidence that a driver actually experiences the car in that way.

The Co-Driving Pleasure Model

The second theoretical output connects the relational account to pleasure and needs:

Co-driving pleasure is a positive emotional experience emerging through the fulfillment of a driver’s needs during situated co-driving activity.

Co-Driving Pleasure Model linking needs, rendering human–vehicle relations, and evaluation of need satisfaction within five contextual factors — The model places needs, enacted relations, and their evaluation inside the context of a driving activity.

A reading of the model moves through needs related to the activity → rendering of human–vehicle relations → evaluation of whether needs are fulfilled. The five influencing factors situate that process.

The HVR framework and pleasure model answer related but different questions. The framework articulates the relation between driver and vehicle; the model considers how such a relation participates in the fulfillment of needs and the emergence of pleasure.

For design, this provides a structured way to frame questions about an intended experience. It is a conceptual model, not an objective pleasure score or a demonstrated causal prediction of a driver’s emotional response.

Design translation: from concepts to things people can use

Three considerations for automotive design

The experiential findings and conceptual work led to three design considerations:

Bring the five contextual factors into design inquiry. Use them to distinguish the situations a concept is intended to support, alongside established safety and usability concerns.
Connect driving habits to experience. Consider both how a system accommodates established practices and how its design may guide the development of new habits.
Design for relations, not only product attributes. Ask how a feature mediates a person’s activity and what relation might develop through repeated, situated use.

Relation modes and the Thing-sona card tool

Relatively stable, sedimented patterns of relating to technology suggested the concept of a relation mode. But a conceptual vocabulary alone was not necessarily an accessible means of discussion.

I developed Thing-sona by connecting relation modes with persona thinking. Its How and Who cards offer a tangible prompt for expressing how a technology is experienced.

Physical Thing-sona cards pairing the How card Dependent and the Who card Secretary — The card grammar combines a relational quality and a perceived role.

Thing-sona contextual board alongside a participant working with the physical cards — The board and cards connect a person's account to a particular technology and situation.

The tool translates the research into something people can interpret, question, and extend. It does not prescribe a single desirable personality for a vehicle.

Explore the dedicated Thing-sona case study for the card system and current digital prototype, or view the repository.

Practitioner dialogue: relevance and limits in design practice

I discussed the research with nine automotive design practitioners, recruited through LinkedIn and personal networks. They worked across technology suppliers, vehicle manufacturers, and design studios.

Process linking design considerations, practitioner talks, feedback, and reflection — Practitioner dialogue examined how the research might enter design practice.

The one-to-one talks followed three steps: a practitioner self-introduction, an introduction to the doctoral research, and discussion of the proposed design considerations.

Three steps of the practitioner talk: self-introduction, research introduction, and design-consideration discussion — The talk structure created space for both understanding the concepts and questioning their practical use.

Responses pointed to the value of considering the full context of a trip and questioning how future experiences could be designed.

Original visual excerpts of responses from design practitioners 4 and 8 about the context of a trip and questioning future possibilities — Selected practitioner responses reproduced from the original case documentation.

At the same time, the discussions exposed three constraints:

Engineering and safety priorities shape the workflow. Practitioner #3 described technical experts as having the strongest influence on cockpit and automated-driving decisions.
Individual habits compete with commercial priorities. Practitioner #2 explained that features for particular driving habits may be deprioritized if their purchasing impact is limited.
Theoretical agreement is not an operational method. A team may appreciate HVR as a principle but still struggle to interpret and apply it consistently across disciplines.

These were not simply endorsements. They made the translation problem more concrete: a relational framework needs usable representations and methods, as well as a persuasive conceptual rationale.

Contribution and continuing questions

The research connects situated accounts of co-driving, three experiential themes, two theoretical tools, and a tangible card-based exploration. Its contribution is a way to investigate and discuss pleasurable human–vehicle relations beyond the assumption that pleasure belongs mainly to manual control.

The next questions concern use in practice: how can multidisciplinary teams work with these concepts, how do relations change over time, and how might contextual differences be accommodated without losing what matters to individual drivers?

The evidence has clear boundaries. The interviews provide qualitative accounts rather than population-level estimates. The second round contributes data-source triangulation; it is not a controlled test of the model. Practitioner discussions indicate perceived relevance and practical tensions, not deployment impact or a measured improvement in experience.

Developing Thing-sona and MCD continues the effort to make relational inquiry accessible through tools, while leaving their effectiveness in real design work open to further examination.

Full doctoral thesis

Peng Lu (2025). Feeling together with the car: towards theorizing co-driving pleasure from a technological mediation perspective. Doctoral dissertation, Politecnico di Milano.

Read the doctoral thesis ↗