Master's Thesis: Automated Data Analysis/Processing of Test Runs

The goal of this master's thesis is to develop an automated data analysis and processing system for extending our in-house software platform Lioness based on measurement and test data. The focus lies on the conception, implementation, and evaluation of a robust condition monitoring system with integrated modules for predictive maintenance. The tasks primarily include:

• Data Analysis and Processing: Identification of relevant sensor data, preprocessing (filtering, normalization, etc.) and feature engineering to characterize the machine state (e.g., RMS values, FFT). 

• Condition Monitoring: Development of algorithms for defining state parameters and threshold values as well as anomaly detection and error classification using statistical methods and machine learning. 

• Predictive Maintenance: Conception and implementation of modules for predicting remaining useful life (RUL) using e.g., regression models or LSTMs. This includes also the fault cause analysis and generation of maintenance recommendations to optimize maintenance strategies. 

• Validation and Evaluation: Evaluation of the developed solutions using suitable performance metrics and practical testing with real or simulated data.