Design and development of an adversarial reinforcement learning approach for situational risk assessment of object identification of a mobile robot.

Topic Area

risk assessment, safety, reinforcement learning, object detection

Task

Knowledge

Contact Person

Manuel Mueller

Security-Bot — Development of a Cybersecurity Analysis with Heuristics and Machine Learning

Topic Area

Security breaches due to errors in cyber-physical systems have far-reaching consequences, including physical damage. One focus of the IAS is to find out how cybersecurity can be reliably and efficiently achieved in such cyber-physical systems.

Task

A tool environment for cybersecurity analysis and grey-box penetration testing will be developed. For this purpose, a TARA is created semi-automatically based on architecture and design properties. From the risk points, e.g. interfaces, SW components, etc., test cases are derived semi-automatically with a bot, which are optimized for efficiency and effectiveness based on KPI.

Knowledge

Lecture SER; Cybersecurity knowledge; Independent work; English language skills: The MT will be written in English.

Contact Person

Simon Kamm

Check-Tool —Environment for Analysis and Assessment of Software

Topic Area

Software must meet complex requirements and standards. Functional safety, cyber security and other quality requirements demand verifiable compliance with specifications, both in terms of technology and development methodology. These must be systematically evaluated.

Task

A professional tool environment for quality assurance checks is to be developed using industry standards. The work is done as an agile project with iterations. First, a shell for such a platform will be modified with refactoring techniques. Optionally, a machine learning component will be developed to provide tips for closing gaps.

Knowledge

Lecture SER C# Programming Independent Work English language skills: The MT will be written in English

Contact Person

Simon Kamm