Digital Twins are virtual representations of physical or cyber-physical objects, systems or processes. Recently, attention to Digital Twins and their use in various industries has increased significantly. The use of Digital Twins offers a variety of benefits, including the ability to monitor in real time, simulate and analyze data, use predictive analytics, remote control and automation, and facilitate lifecycle management. The development of Digital Twins is based on the integration of models and data coming from a variety of sources, including sensors, Internet of Things (IoT) devices and other data streams.

Model-driven development is considered an effective technique for overcoming the challenges associated with developing IoT applications. The automatic generation of code resulting from model transformations increases productivity and ensures consistency through automation.

Existing model-based approaches to IoT application development focus on the software and system perspective. We have recognized the need to integrate organizational aspects into the development of IoT applications. There is no approach that is tailored to the specific needs of small and medium-sized enterprises (SMEs). The requirements of SMEs for the development of IoT-based digital twins differ from those of larger companies. There are significant differences in terms of available resources, IT structure, existing knowledge and general aspects such as the SME's strategy or culture.

The aim of this doctoral thesis is to develop a method that focuses on the integration of organizational aspects in the development process as well as the specific needs of SMEs in the development of IoT-based digital twins. The method components include guidelines, steps and the required methods for application. We have developed a tool that supports the entire development process without the need for specialized IT knowledge at the application level. This should help to (i) incorporate the necessary domain knowledge, (ii) support the construction of digital twin models and (iii) create the actual digital twin with its functionalities.

Energy management refers to the systematic planning, control, and monitoring of energy use in companies with the aim of reducing energy consumption and costs while contributing to climate protection. Energy management systems (EnMS) provide a structured organizational and IT framework that supports companies in collecting and evaluating energy-related data and continuously deriving improvement measures. The introduction of such systems is becoming increasingly important, particularly against the backdrop of rising energy prices and regulatory requirements.

The doctoral project focuses on small and medium-sized enterprises (SMEs), which have high savings potential but are often confronted with specific challenges when introducing energy management systems. These include limited human and financial resources, a lack of specialized knowledge, and the poor fit of existing, often highly standard-oriented solutions. As part of the doctoral project, the needs, requirements, and framework conditions of SMEs are surveyed in a practical manner and directly integrated into the scientific investigation.
The empirical basis of the doctoral project consists of qualitative interviews, standardized surveys, and in-depth case studies in companies. This combination of methods is intended to identify both individual experiences and overarching patterns and influencing factors in the introduction and use of energy management systems.

The aim of the doctoral project is to develop a practical tool or framework that provides targeted support to companies – especially SMEs – in the introduction and implementation of energy management systems. This tool should be scientifically sound, but at the same time adapted to the real requirements of business practice. One possible design is the development of a reference model that serves as a guide and decision-making aid for companies and facilitates the successful implementation of energy management systems in the long term.

Generative language models are becoming increasingly important as a basic technology for assistance and decision support systems in organizational contexts. Despite their high performance, however, challenges remain with regard to systematically ensuring quality, reliability, and domain suitability. Particularly when used in knowledge-intensive application scenarios, a structured quality management approach is necessary to ensure trust, traceability, and acceptance of the generated results.

The goal of this doctoral project is to develop a domain-specific quality management process for the use of generative language models. Based on a systematic analysis of existing evaluation and quality approaches, a proprietary quality model will be developed that integrates technical, subject-specific, and user-related quality dimensions. The approach combines automatic and human evaluation mechanisms and enables continuous assessment of model quality throughout its entire life cycle.

The doctoral project is closely embedded in an ongoing research project to develop an LLM-based digital coach. The developed quality management process forms a central basis for ensuring the effectiveness and reliability of the coach in domain-specific application scenarios. The work thus contributes to the responsible use of generative language models and the further development of quality management concepts in business informatics.