The International Space Station ISS is a complex technological system. In the event of an error, e.g. a disturbance in the life support system, a quick and targeted identification of the cause of the error is necessary. At the moment, however, it is an increasing challenge for the experts in the ground station to understand the data from 20,000 sensors and the complex interactions in the station. In this project, data-based methods of machine learning and knowledge-based methods of symbolic artificial intelligence are therefore combined in order to develop an assistance system to support the operator.
The aim of this project is to develop algorithms and to implement the appropriate practical software solution that provides the following functionalities:
• Automatic AI-driven detection of anomalies in different subsystems (the ISS) by analyzing streaming data
• Sending of notifications to the expert team in Bremen and to the control center about the occurrence of the anomalies
• Automatic and AI-driven diagnosis (root cause analysis) for detected anomalies
• Provision of recommendations for action and repair instructions to remedy the anomalies
for the responsible engineers
• In the case of critical, time-critical errors, initiation of measures for automatic repair
Duration: November 1, 2020 to December 31, 2024
The project is funded by the Federal Ministry of Defense (BMVg) in the DTEC funding program.
Letzte Änderung: 24. February 2021