Information systems and decision support for sustainable urban transport, energy systems, and emerging methods in research and practice

The dissertation follows a multi-level research approach based on design science principles and integrates different methods to generate insights and develop decision support functions to promote sustainable urban development in the fields of mobility, energy and logistics. The approaches include spatial data analysis, optimization modelling, machine learning, text mining and the development of decision support systems (DSS). Some of the methods used are: Spatial data analysis and geographic information systems (GIS): These methods are used to analyze e-scooter usage patterns, determine the optimal placement of wind turbines, and study the factors that influence the adoption of residential solar systems. Optimization modeling: The dissertation presents optimization models to support strategic and operational decisions in different contexts. For example, optimization models are used to optimize the logistics chain for urban deliveries, support decision making for operators of aging wind turbines, and optimize a smart logistics concept for e-food operations. Data mining and supervised and unsupervised learning: Machine learning and data mining techniques are used to gain insights from unstructured data. For example, web content mining, natural language preocessing, and topic modeling is used to analyze causes of escooter accidents. Semantic analysis and topic modeling is applied to social media data and scientific articles to understand the perception and discourse on large language models such as ChatGPT. Also clustering and regression methods are applied on spatial data and taxonomy-based data. Delphi study: A Delphi study is conducted to gain expert knowledge and validate simulation results in the context of sustainable logistics scenarios. The core findings can be divided into three main thematic areas: Sustainable urban transportation and mobility: in the field of urban logistics, a decision support system (DSS) is developed to help decision makers evaluate and compare different urban logistics scenarios to enable informed and environmentally conscious decisions in the design of urban and rural logistics networks. This DSS quantifies the trade-offs between cost efficiency and sustainability, helping practitioners to manage the complexity of green logistics in urban and rural environments. • Evidence shows that establishing shared freight consolidation centers outside city centers with micro-depots near city cores can reduce truck traffic and fuel consumption. • Shared micromobility research shows spatiotemporal e-scooter usage patterns, accident causes, and factors affecting shared e-scooter, bicycle, and e-moped usage. Sustainable energy systems: The dissertation focuses on the management of wind turbines and residential solar systems. • GIS frameworks are created to support decision making for aging wind turbines, analyze economic viability and regulatory challenges. The results provide important insights for the end-of-funding analysis of wind power plants, whereby a prompt repowering one to two years after the end of a subsidy can generally be suggested. The factors influencing the adoption and pricing of residential solar systems in Germany are examined, revealing significant spatial solar system price differences in Germany and a significant positive relationship is between green voters and plug-in solar system adoption. Emerging technologies and methods in the IS domain: The dissertation also examines the applications of advanced methods and technologies in IS research and practice. • Taxonomy-based data clustering methods are investigated to gain insights providing guidelines for their application. • The social and scientific perception of large language models, especially generative, pre-trained, transformer-based chatbots, is analyzed uncovering potential areas for improvement providing guidance for future research and development in the field of conversational artificial intelligence. Encountered limitations highlight the importance of transparency in the assumptions, scalability, integration with existing IT infrastructures, and evaluation of DSS.