On June 5, 2025, Niklas Götz successfully defended his master's thesis on “Development and Integration of a Web Application for the Monitoring and Analysis of AI-Based building blocks”. Niklas is a master's student in business informatics and was supervised by Nikola Ivanovic and Prof. Kurt Sandkuhl (both employees of the Chair of Business Informatics).
Abstract
This thesis presents the development and integration of a web application for monitoring and analyzing AI-based building blocks in HVAC systems. Building on an existing digital twin, the aim is to create a modular, scalable infrastructure that enables systematic evaluation of sensor data, anomaly detection, and user-friendly visualization. A dedicated AI component was developed, utilizing machine learning techniques—particularly autoencoders and isolation forests—for anomaly detection in time series data. The results are visualized via a custom-built web application based on a Model-View-Controller (MVC) architecture. The work combines concepts from modular software design, design patterns, and AI system architecture to ensure reusability, scalability, and transparency.