Research in real-time dependable systems and risk assessment
My objective is to contribute to the techniques, tools and computing platforms that will make it possible to build provably safe systems in a time and cost efficient manner, with applications to:
- Automotive and aerospace embedded systems: automated synthesis of E/E architecture (design-space exploration, machine learning) & timing and dependability verification
- Risk assessment: quantify risks using probabilistic models built from historical data
Here is a list of my publications (most are available for download) and slides from recent talks. You can find here a short biography and a list of (academic) professional activities.
Check-out CPAL - our proposal for the next generation low-code language to design, simulate and execute embedded systems.
- E/E architecture design automation:
- A Model-Based Systems Engineering framework with DSE for E/E architectures in automated driving vehicles, with Robert Bosch.
- Graph neural networks to speed-up the verification of Ethernet TSN networks in design-space exploration - study available here as well as a follow-up work.
- “QoS-Predictable SOA on TSN: Insights from a Case-Study”, with Renault.
- “Towards Computer-Aided, Iterative TSN-and Ethernet based EE Architecture Design”, with BMW.
- “Early-stage Bottleneck Identification and Removal in TSN Networks“, with Volvo.
- “Practical Use-cases for Ethernet Redundancy”, with NXP.
- "Multi-source software on multicore automotive ECUs - Combining runnable sequencing with task scheduling", with PSA.
- Communication networks for dependable systems:
- Timing QoS protocols on top of Ethernet TSN (slides)
- Scheduling frames with offsets provides a major performance boost on CAN (slides)
- Configuration of FlexRay networks (slides)
- Fine Tuning MAC Level Protocols for Optimized Real-Time QoS
- Optimal configuration of TDMA / TTP/C networks - (slides)
- Probabilistic analysis of CAN fault-confinement mechanisms
- Patents on communication networks: FR2976432 - FR2976434 - FR2976435.
- Model-Driven Engineering for embedded systems:
- Low-power scheduling:
- Optimal CPU frequencies for energy consumption, the case of FIFO tasks and the general case.
- Stochastic model of battery discharge time
- Multi-processor low-power scheduling
- Financial engineering