Towards White-Box Optimization in Self-Adaptive Systems
| Typ | Masterarbeit | |
|---|---|---|
| Aushang | Towards White-Box Optimization in Self-Adaptive Systems.pdf | |
| Betreuer | Wenden Sie sich bei Interesse oder Fragen bitte an: Ehsan Elahi (E-Mail: ehsan.elahi@kit.edu, Telefon: +49-721-608-41630) |
Motivation
Since Cyber-Physical Systems require continuous and uninterrupted operation, self-adaptive systems (SASs) have become essential for managing the uncertainty in the environment and responding to unforeseen changes at runtime. Their ability to adapt with minimal human intervention makes them useful, particularly in safety-critical domains. However, current optimization approaches used within SASs often function as black boxes, limiting interpretability, thus making the computation of adaptation decisions inefficient. This gap highlights the need for white-box optimization methods that provide transparency, support reasoning about adaptation decisions, and operate efficiently under uncertainty. Addressing this need requires a systematic investigation and the development of a white-box optimization methodology within SASs.
Tasks
- Investigate the existing work on the white-box optimization in SASs
- Develop an algorithm or methodology for white-box optimization in the context of a SAS, such as robotics
- Validate and analyse the proposed methodology to identify its improvement in terms of efficiency and effectiveness.
Benefits
- Working with cutting-edge and innovative technologies
- Excellent working environment with regular supervision meetings.
- Possibility to co-author or contribute to a scientific publication (depending on results)