With the arrival of automated vehicles (AVs) on our streets virtually around the corner, this thesis explores advances in automated driving technology with a focus on ethical decision making in dilemmatic traf- fic situations. In a total of five publications, we take a multi-facetted approach to analyse and address the core challenges related to auto- mated ethical decision making in AVs. In publications one through three, we conduct a series of immersive virtual reality studies to analyze human behavior in traffic dilemmas, explore mathematical approaches to model the decision making process, investigate how the assessment methodology can affect moral judgment, and discuss the implications of these studies for algorithmic decision making in the real-world. In publication number four, we provide a comprehensive summary of the status quo of AV technology and legislation with regard to automated ethical decision making. Here, we discuss when and why ethical deci- sion making systems become necessary in AVs, review existing guide- lines for the behavior of AVs in dilemma situations, and compile a set of 10 demands and open questions that need to be addressed in the pursuit of a framework for ethical decision making in AVs. Finally, the basis for automated ethical decision making in AVs will be provided by accurate assessments of the immediate environment of the car. The pri- mary technology used to provide the required information processing of camera and LiDAR images in AVs is machine learning, and in particular deep learning. In publication five, we propose a form of adaptive acti- vation functions, addressing a central element of deep neural networks, which could, for instance, lead to increased detection rates of relevant objects, and thus help to provide a more accurate assessment of the AVs environment. Overall, this thesis provides a structured and compre- hensive overview of the state of the art in ethical decision making for AVs. It includes important implications for the design of decision mak- ing algorithms in practice, and concisely outlines the central remaining challenges on the road to a safe, fair and successful introduction of fully automated vehicles into the market.

In this thesis, a conceptual framework and related methodological approaches for complex system design are developed and tested. The approaches are based on insights from the fields of Systems Engineering (SE) and System-of-Systems Engineering (SoSE), as well as Natural Resources Management (NRM). The focus of this thesis is on: 1) the development of the System-of-Systems Design Framework “FRESCO”, 2) the development of a methodological framework for participatory systems design, 3) the application of the framework in two case studies, and 4) the development of an evaluation scheme to qualitatively measure the effectiveness of the methodological framework. The overall objectives of this doctoral dissertation are to highlight synergies between SE and NRM and to develop a methodological framework for designing decision making processes in a human-nature-technology context. The complexity of coupled and complex adaptive systems (CAS) such as the Water-Energy-Food Nexus (WEF-Nexus) and sustainability strategies, influences the design of decision-making processes and strategy building. Integrated process design which is promoted by the developed frameworks can assist in such tasks on an urban, regional, and national level.

A recent meta-analysis of Anderson and colleagues (2010) shows that violent behavior in computer games promotes violent behavior in real-life and inhibits prosocial behavior. A couple of studies conducted by Greitemeyer and Osswald (2010) lead to the conclusion that helping behavior in computer games furthers helping behavior in real-life. There exist no studies examining the combined effect of violence and helping in computer games, although this combination is typical for violent video games (Anderson et al., 2010). In violent RPGs, a lot of tasks consist of helping someone by using violence. The present study addresses this issue and bridges the current empirical gap by investigating if violent emergency assistance furthers helping behavior and/or violent behavior in real-life. To accomplish that, the role-playing game “Oblivion” was modified to create four different experimental conditions: (1) violent emergency assistance, (2) killing, (3) helping, and (4) treasure hunting. Comparing these conditions, violent emergency assistance seemingly reduces helping behavior in real-life and at the same time furthers violent behavior. The results are in unison with the moral management model (Hartmann & Vorderer, 2010; Hartmann, in press), which is based on Banduras Theory of Moral Disengagement (Bandura, 2002).

A fundamental change in societal values and economic structures is required to address increasing pressures on ecosystems and natural resources. Transition research has developed in the last decades to analyze the co-dynamics of technological, institutional, social and economic elements in the provision of key functions such as energy, water and food supply. This doctoral dissertation provides conceptual and methodological contributions to the pro-active governance of sustainability transitions. Three research gaps are identified that are addressed in this dissertation. First, a comprehensive conceptualization of learning in sustainability transitions is currently missing that comprises learning at multiple societal levels (ranging from individuals to policy-actors). Learning concepts are often not explicitly discussed in transition research even though learning is considered as fundamental for innovation processes, niche formation and development as well as breakthrough and diffusion of innovations. Second, methods for the analysis and design of transition governance processes are lacking that specify case-specific intervention points and roles of actors in the implementation of innovations. Third, participatory modeling approaches are only applied to a limited extent in transition research despite a high potential for supporting communication and learning. The conceptualization of multi-level learning developed in this doctoral research conceptualizes learning at different societal levels as specific learning contexts ranging from individual and group contexts to organizational and policy contexts. The conceptual framework further differentiates between learning processes, intensity, objects, outcomes, subjects and factors, allowing for a more detailed analysis of learning within and across learning contexts. Thus, learning contexts can be linked by processes that involve actors from different learning contexts (e.g., community groups and policy-makers), as well as exchanges of physical aspects, institutions and knowledge (in the form of ‘learning factors’). This research has also provided a classification of model uses in transition research that supports a purposeful discussion of the opportunities of modeling and promising future research directions. The methodology developed in this doctoral research aims at the analysis and design of transition governance processes by specifying the various opportunities to contribute to sustainability transitions through purposeful action at different societal levels, as well as related roles of stakeholders in implementing such processes of change. The methodology combines different streams of previous research: 1) a participatory modeling approach to identify problem perceptions, case-specific sustainability innovations as well as related implementation barriers, drivers and responsibilities; 2) a systematic review to identify supportive and impeding learning factors from the general literature that can complement case-specific factors; and 3) a method for the analysis and design of case-specific transition governance processes. Three case studies in Canada (topic: sustainable food systems), Cyprus (water-energy-food nexus) and Germany (sustainable heating supply) have been selected to test and iteratively develop the methodology described above. The results for each case study reveal that there are learning objects (i.e., learning requirements) in all learning contexts, which underscores the importance of multi-level learning in sustainability transitions, ranging from the individual to the group, organizational and policy levels. Actors have various opportunities to actively facilitate societal transformations towards sustainable development either directly through actions at their particular societal levels (i.e., context-internal learning) or indirectly through actions that influence learning at other societal levels. In fact, most of the learning factors require cooperation across learning contexts during the implementation process. The comparing of learning factors across case studies underline the importance of several factor categories, such as ‘physical a ‘disturbance or crisis’, ‘information and knowledge’. Of the 206 factors identified by stakeholders, 40 factors are case-specific and not contained in the general, review-based factor list. This underscores the value of participatory research, as general, top-down analyses might have overlooked these case-specific factors. The methodology presented in this dissertation allows for the identification and analysis of case-specific intervention points for sustainability transitions at multiple societal levels. The methodology furthermore permits the analysis of interplay between individual, group, organizational and policy actions, which is a first step towards their coordination. The focus on sustainability innovations links the broad topic of sustainability transitions to a set of opportunities for practical interventions and overcoming their implementation barriers. The methodology presented allows for the analysis and design of these interlinkages between learning contexts. While the methodology cannot provide any ‘silver bullets’ for inducing sustainability transitions, it is flexible enough to identify an appropriate abstraction level for analyzing and designing transition governance processes. The methodology developed in this doctoral research also provides several contributions for the development of participatory modeling methods in transition research. Thus, the participatory method supports an integrated analysis of barriers and drivers of sustainability innovations, and allows application in practice and education. The concepts and methods developed in this research project allow for reflection on transition governance processes from a systemic viewpoint. Experiences in the case studies underline the applicability of the concepts and methods developed for the analysis of case-specific transition governance processes. Despite substantial differences in the geographic location, culture and topics addressed, all case studies include promising sustainability innovations and the engagement of multiple actors in their implementation. The diversity and multitude of initiatives in the case study regions provides an optimistic outlook on future opportunities for large-scale sustainability transitions.