DC2: Ensuring Runtime Safety: A Digital Twin Approach for ML-Based Controllers
Project: Ensuring Runtime Safety: A Digital Twin Approach for ML-Based Controllers (WP1)
Host institution: FGH-IESE (Germany)
Supervisor: Dr. R. Adler (FGH-IESE, Germany)
Co-supervisor(s): Davy Pissoort (KU Leuven), G. Lunzenfichter (Medisanté)
Objectives:
Create digital twin technology for individuals using assistive health technolgy devices to enhance real-time safety assurance of machine learning-based controllers; Assess the effectiveness of the developed methodology through a comprehensive evaluation in a single-use case study
Expected Results:
Establishment of a universal framework seamlessly integrating real-time safety assurance for machine learning-based controllers through digital twin technology; Demonstration of the digital twin technology for assisted individuals in simulation, focusing on a selected case study.
Planned secondment(s):
- KU Leuven (3 months, M22-M24) Study how to adapt and incorporate the run-time safety monitor framework with digital twin technology (3 months, M14-M16):, with Davy Pissort (KPI: joint conference paper)
- Medisanté (3 months, M38-M40) Application and evaluation – with respect to scientific criteria as well as practical applicability – of the platform in an industrial setting: with G. Lunzenfichter (KPI: joint journal paper)
Enrolment in Doctoral degree: Doctoral School of Technische Universität Kaiserslautern (Germany)
Required profile: Software Engineering
Desirable skills/interests: Safety, Digital Twins, Data Science