The typology is based on concepts from transition research and TA. A core tenet of transition research is the conceptualization of infrastructures as socio-technical systems, which are shaped by the co-evolutionary interplay of technical and non-technical factors. The regime is usually understood as the semi-stable institutional core of a socio-technical system and is strongly influenced by so-called normative and cognitive institutions (Fuenfschilling and Truffer 2014), which are often not directly visible and empirically accessible (Geels 2011). These are, for example, taken-for-granted assumptions, routines, attitudes, expectations or broader social values (Rip and Kemp 1998). Together with ‘visible’ elements, such as technologies, actors, and regulations, this institutional regime configuration co-determines the future development direction of a socio-technical system. A socio-technical regime does not necessarily have to resemble a homogeneous institutional configuration, but can be characterized by different, sometimes competing ideas about what is possible and what is a desirable direction of change. For example, in the mobility system we can observe that there are very different views on whether the private car should retain or lose its dominant position in future mobility.

The concept of directionality is used to indicate that transformative change in a system or sector has a specific direction, while other developmental directions are in principle also possible (Stirling 2009; Weber and Rohracher 2012). In the literature on innovation and socio-technical transitions, the concept is usually linked to the idea of promoting not just any innovation trajectory, but those that contribute to a desirable direction of transformative change and help mitigate societal challenges (Parks 2022). It is repeatedly emphasized that in some cases, especially in the case of wicked problems, such shared visions of desirable futures cannot be achieved without challenges and conflicts (Schlaile et al. 2017; Wanzenböck et al. 2020). As already mentioned, different ideas can exist in a regime about which direction of change is desirable. While the term directionality on the one hand refers to the direction of an entire sector, on the other hand, directionality can imply different development conditions for different technologies. In this sense, Yap and Truffer (2019, p. 1032) understand directionality from the perspective of innovations as being “tightly associated with the specific form of the selection environment to which the technology is exposed”. Drawing on this understanding, in this paper, I want to highlight that different innovations can be categorized in different ways with respect to directionality (Schippl 2024). Different modes of directionality are distinguished on the basis of two dimensions: The first dimension refers to the characteristics of the technology itself, particularly its possible range of applications. The second dimension refers to the specific selection environment of the technology, primarily to its normative contextual conditions, such as the homogeneity or heterogeneity of societal expectations, preferences or attitudes towards the potential applications of the new technology.

In the context of TA, the typology refers to the classification of five different epistemological insights or types of knowledge introduced by Armin Grunwald (2020). The governance of innovation and transition processes requires knowledge of different epistemic character (Grunwald 2018). Obviously, there is a need for what is referred to here as ‘empirical system knowledge’ in order to achieve a good understanding of the conditions and causal chains in the current system. However, not all of the relevant factors and interrelationships are directly ‘visible’ and empirically accessible (Schippl 2024). For example, research into the configurations and dynamics of cognitive-normative institutions requires more interpretative, hermeneutic approaches (hermeneutic knowledge). Obviously, there is also a need to understand plausible future development paths of an innovation and potential consequences that lie in the future and therefore are again not empirically accessible (future knowledge). In addition, knowledge of the prevailing target systems (e.g. phase-out of internal combustion engines by 2035) and the underlying reasoning patterns is required. This category of normative knowledge aims primarily at ‘direct’ orientation in the sense of identifying desirable or undesirable directions of development with more or less specific target settings. The fifth category that can be distinguished is instrumental knowledge, which is primarily concerned with the concrete and practical implementation of measures, such as technical standards or legal rules. As will be shown, different types of innovation place different demands on the knowledge categories from which specific strategies or tactics for (long-term) governance can be derived.

Figure 1 depicts different modes of directionality based on the abovementioned two dimensions. The y‑axis shows the scope of possible future applications of the technology in terms of the primary function of the socio-technical system, the provision of mobility. For some technologies this range is rather narrow, while others have a wide range since they can be used for different purposes. The x‑axis indicates whether the technology meets with a normatively more heterogenous or a more homogenous regime configuration in the selection environment. As will be shown below, the normative homogeneity of the mobility regime is significantly higher for EV than for AV. On the basis of these two dimensions, the following three ideal modes of directionality can be identified (Grunwald 2014; Schippl 2024).

In the case of mode-1-directionality, the respective regime configuration is rather homogeneous and the application of the new technology is quite obvious. This is the case, at least so far, for battery electric propulsion in private cars, which will serve as an example here. The scope of future applications is rather narrow, at least when it comes to car-based mobility: An internal combustion engine (ICE) is replaced by electric propulsion, the car remains a car. The question is how to ensure a rapid uptake of electric cars, while car-based mobility as such is not affected by the new technology.

In the political arena in Germany e‑mobility is predominantly seen as a desirable development and is supported by corresponding regulations.

Looking at the relevant institutional configurations in the selection environment (x-axis in figure 1), it can be seen that there are very clear objectives for electric mobility. In the political arena in Germany (and in the EU; e.g. the ban for ICE in 2035 - even if the date may be postponed), e‑mobility is predominantly seen as a desirable development and is supported by corresponding regulations. The German government has set a political target of 15 million electric cars in the car fleet by 2030. Yet, there are challenges and reasons for concerns which will be of relevance from a long-term perspective. The following issues are among the most relevant:

However, technical progress and regulations can solve these issues. For this, future knowledge, system knowledge and instrumental knowledge are primarily crucial. Future knowledge can be generated, for example, through model-based exploratory scenarios to anticipate future renewable electricity demand and charging point locations. System knowledge includes research on new batterie materials or on understanding of the variability of charging patterns, for example based on user surveys, in order to calculate the impact of these variations on the energy system. Instrumental knowledge includes developing regulations or incentives for the installation of charging stations, or the adjustment of charging behavior to the availability of electricity from renewable sources.

AV may exemplify mode‑2. Automated driving has great transformative potential for the future of mobility and society (Cohen and Cavoli 2019). AVs can be used in conventional cars to relieve drivers and enable new services like driverless robo-taxis or shuttles. Various long-term path-dependencies are possible (Schippl et al. 2022). It is currently uncertain whether AVs will create a sustainable, climate-friendly mobility system by providing flexible, accessible and affordable public transport options, thereby significantly reducing the use and ownership of private cars, or whether they will make private cars more attractive, ultimately leading to increased emissions, land consumption and urban sprawl (Lyons 2022). Nevertheless, AVs must adapt to existing infrastructures and mobility needs (Fraedrich et al. 2015), which have co-evolved over decades and cannot easily be changed. The scope for future development is therefore bounded. Despite AV’s potential to change the mobility system in different ways, plausible future developments are fairly well understood (Schippl 2024). From a long-term governance perspective, the main challenge is the prevailing normative heterogeneity in the mobility sector, in particular when it comes to the role and relevance of the private car. To counter unsustainable path dependencies, clear visions and objectives for the mobility system as a whole are needed, but at present these are at best fragmented and controversial. Many observers point to this lack of strategic clarity and direction (Horn et al. 2018; OECD 2018). Drexler et al. (2022) conclude that there is no common understanding among actors in the German mobility sector of the problems and solutions for promoting a mobility transition. As a result, there are no clear overarching goals to guide the development of AV. AV interact with a selection environment that is open to both sustainable and unsustainable path dependencies (Schippl 2024).

The German government recently published a strategy for autonomous driving on the roads (BMDV 2024), focusing on promoting AV in public transport as key driver for AV development. This supports sustainable path-dependency. However, a robust long-term perspective requires linking the strategy to a broader vision of future mobility in Germany, which would provide a target system or at least a coherent orientation framework for long-term governance. Austria provides an example of how such a lack of normative knowledge can be addressed: A few years ago, a mobility master plan was adopted with a clear vision and concrete targets for the model split in 2040 (BMK 2021). Furthermore, in the case of mode-2-directionality, hermeneutic knowledge is needed, especially to better understand the underlying reasons for normative heterogeneity. Again, there are studies that could provide starting points for future research. Groth et al. (2023), for example, examined how everyday mobility is portrayed in popular crime series and found that the current controversy about the future role of the car is also reflected in the media, with car-centered storylines dominating. Mögele and Rau (2020) show that a culturally sensitive approach can reveal differences in mobility culture that have an impact on the development of sustainable mobility.

A current example of mode‑3 directionality in mobility is blockchain technology. The technology is evolving, with many applications being discussed and tested in various sectors. Some align with current mobility trends like mobility-on-demand, autonomous driving, and electrification (Karger et al. 2021). Blockchain could lead to significant changes in the mobility system (Gösele and Sandner 2019) such as simplifying carpooling and private vehicle rentals, increasing efficiency, reducing cost, and mitigating market power concentration. However, blockchain aims to replace controlling intermediaries such as commercial institutions with decentralized peer-to-peer networks. In the mobility sector, it has to be considered whether a self-regulating system could develop that evades public planning and control or is possibly controlled by a market-dominating tech company. Blockchain’s immutable recording of transactions can pose privacy problems and data protection issues. Potential developments and their effects are not yet well understood.

To better understand the range of possible developments, future knowledge is needed. It is advisable to use structurally open methods (Schippl and Fleischer 2012) such as brainstorming, open space, world café, etc., which enable a deliberate generation of new ideas. Furthermore, it is important to better understand the ideas, fears and expectations regarding the possible application contexts of the innovation and their background (Grunwald 2014). This requires interpretive, hermeneutic approaches. Such expectations have a direct or at least indirect impact on the design of the technology – otherwise it would be a case of technological determinism. It is by means of such methodological approaches that the basis for normative evaluations can be created; the innovation (or specific fields of application) may then become mode‑2 or mode‑1 directionality.