Managing Conflicting Interests in Socio-technical Energy Systems – How to Identify and Mitigate Intra-actor Interests as Risk Factors Astrid Nieße1[0000-0003-1881-9172], Frank Eggert3[0000-0002-5797-4741] and Sebastian Lehnhoff2[0000-0003-2340-6807] 1 Leibniz University Hannover, Germany 2 TU Braunschweig, Germany 3 University of Oldenburg, Germany niesse@ei.uni-hannover.de Abstract. Global networks like energy grids, transportation networks or financial IT-infrastructure are crucial for the wealth of modern societies. Reliable and re- silient control of these infrastructures thus has gained much attention in the last years. Typical approaches to ensure stable operation of these infrastructures fol- low two contradictory paradigms, i.e. complexity reducing and complexity in- creasing measures. Whereas the first are supposed to encapsulate interdependen- cies and decision-making processes and typically reduce transparency as a side- effect, the latter strengthen the role of the human actor in these systems by in- creasing transparency to allow for well-informed decision-making. In this paper, we will discuss these two paradigms and show why intra-actor con- flicts arise from adding both complexity and reducing transparency at the same time. We will outline a research agenda to model the effect of these conflicts using the example of energy systems and current transparency-enhancing tech- nologies like e.g. distributed ledger technology. Keywords: smart grids, socio-technical energy systems, intra-actor conflicts 1 Summary Global networks like energy grids, transportation networks or financial IT- infrastructure are crucial for the wealth of modern societies. Reliable and resilient con- trol of these infrastructures thus has gained much attendance in the last years [1]. Typ- ical approaches to ensure stable operation of these infrastructures follow two contra- dictory paradigms: Complexity reducing measures are supposed to encapsulate interdependencies and decision-making processes. With energy systems being complex cyber-physical sys- tems and setting up a critical infrastructure of modern societies, dependencies and ef- fects of decisions can only hardly be evaluated [2]. Thus, modern ICT systems try to hide relevant information in such a way that decisions are supported best while reducing 41 potential non-relevant interdependencies. These measures typically reduce transpar- ency as a side-effect. Two technologies largely support the trend to transparency and complexity reduc- tion:  Machine Learning: The availability of data has largely supported the role of ma- chine learning and other data-driven AI-based technologies like neural networks. Supported by affordable multi-core computing hardware, automated learning has taken the role of expert systems without providing additional methods for inter- pretability [3]  Autonomous agent-based systems: Autonomous systems, often based on soft- ware agents and multi-agent systems, pave the way to a decentralized and adaptive control of both processes and systems [4]. Given the high degree of interconnect- edness of technical systems as e.g. used in IoT (Internet-of-Things)-applications, decentralized decisions of technical components reduce the overall understanda- bility and global interpretability of decisions. Complexity increasing measures on the other hand, strengthen the role of the human actor in these systems by increasing transparency to allow for well-informed decision- making. In energy systems, this is of crucial importance, as security, stability and ro- bustness of the system operation largely depends on (human) decision processes. There- fore, these decision processes should be supported by giving transparent access to all relevant data. Many trends in current technologies support the trend to complexity in- crease and transparency, e.g. distributed transaction systems: especially distributed concepts like distributed ledger realizations, allow to track either contract-relevant or even safety-critical processes. By adding data access concepts well-known from cryp- tography, even privacy-relevant or safety-critical data can be stored in a cooperative and secure manner, thus allowing to give access to relevant human actors for decision processes [5]. In this paper, we will discuss these two paradigms and show why intra-actor conflicts arise from adding both complexity and reducing transparency at the same time. We will outline a research agenda to model the effect of these conflicts using the example of energy systems, autonomous control systems and current transparency-enhancing tech- nologies like e.g. distributed ledger technology. We will focus on the role of the human actor in this system and the effects of possible intra-actor conflicts: Human actors base their decisions on information regarding the effects of possible actions (and alterna- tives). Consequently, increasing / reducing transparency will lead to different decisions: Adding automated process steps encapsulates these and enhances system controllabil- ity, but reduces interpretability. As a consequence, decision processes may be encum- bered. Adding transparency on the other hand will lead to a more complex decision setup. Thus, a dynamic, customizable trade-off is needed in the IT-enhanced control of cyber-physical systems like energy systems. Up to now, no engineering process has been presented to support the development of this kind of systems. The research agenda presented in this work addresses current approaches and open issues regarding an abstract model of decision conflicts, a dynamic model to evaluate the effect of transparency changes during runtime, and metrics to evaluate degrees of autonomy and transparency, using the context of energy systems. 42 References 1. Uday, P., & Marais, K. (2011). Designing Resilient Systems-of-Systems: A Survey of Met- rics, Methods, and Challenges. Systems Engineering, 18(5), 491–510. https://doi.org/10.1002/sys 2. Chopade, P., & Bikdash, M. (2011). Critical infrastructure interdependency modeling: Us- ing graph models to assess the vulnerability of smart power grid and SCADA networks. 2011 8th International Conference & Expo on Emerging Technologies for a Smarter World, 1–6. https://doi.org/10.1109/CEWIT.2011.6135885 3. Lipton, Z. S. (2018): The Mythos of Model Interpretability. Communications of the ACM, 61(10), 36–43. 4. Ramchurn, S. D., Vytelingum, P., Rogers, A., & Jennings, N. R. (2012). Putting the “ Smarts ” into the Smart Grid : a Grand challenge for artificial intelligence. Communications of the ACM, 55(4), 86–97. https://doi.org/10.1145/2133806.2133825 5. Unterweger, A. Knirsch, F., Leixnering, C., and Engel, D. (2018). Lessons learned from Implementing a Privacy-Preserving Smart Contract in Ethereum. In: 9th IFIP International Conference on New Technologies, Mobility and Security (NMTS), Paris, France. 43