Runtime Assurance for Autonomous Vehicles (Wintersemester 2023/2024)

Dozent: Prof. Dr. Holger Giese (Systemanalyse und Modellierung) , Dr. Maria Maximova (Systemanalyse und Modellierung) , Dr. Sven Schneider (Systemanalyse und Modellierung) , He Xu (Systemanalyse und Modellierung) , Sona Ghahremani (Systemanalyse und Modellierung)

Allgemeine Information

• Semesterwochenstunden: 4
• ECTS: 6
• Benotet: Ja
• Einschreibefrist: 01.10.2023 - 31.10.2023
• Lehrform: Projektseminar
• Belegungsart: Wahlpflichtmodul
• Lehrsprache: Englisch

Studiengänge, Modulgruppen & Module

Beschreibung

Abstract:

The increasing complexity and criticality of autonomous driving systems necessitate advanced methods for ensuring their safety and reliability. This seminar provides participants with a multi-faceted approach to understanding and ensuring the safety of autonomous vehicles through graph transformation techniques. Beginning with a foundation in literature, participants are introduced to crucial papers in the domains of graph transformation, traffic simulation, and reinforcement learning for self-driving. The seminar then delves into practical applications, where students utilize the SUMO tool for traffic network simulation, extracting and transforming data into the GROOVE format for dynamic verification. The results of this verification are integrated back into SUMO for a holistic view. In an advanced phase, students combine SUMO and Carla to develop a rudimentary self-driving car while simultaneously employing the GROOVE system for real-time graph transformation, enabling them to predict and monitor potential unsafe operations of the autonomous vehicle.

Goal of the Seminar:

1. Theoretical Exploration: Equip participants with foundational knowledge from seminal papers on graph transformation, traffic simulation, and reinforcement learning for autonomous vehicles.

2. Practical Application with SUMO: Introduce the traffic network simulation using the SUMO tool, and create, understand, and extract real-time data from these simulations.

3. Data Transformation to GROOVE: Process the extracted data, and transform it into the GROOVE format suitable for dynamic verification.

4. Dynamic Verification: Utilize the GROOVE system to construct a graph transformation system, and verify the data dynamically and understand its implications for autonomous vehicle safety.

Advanced Goal:

1. Integration with Carla: Provide hands-on experience in autonomous vehicle development by integrating the capabilities of SUMO and Carla.

2. Real-time Monitoring and Safety Predictions: Combine GROOVE, SUMO and Carla for real-time formal verification, enabling continuous monitoring of the autonomous model and preemptive identification of potential unsafe operations or situation.

Lern- und Lehrformen

The course is a project seminar, which has an introductory phase comprising initial short lectures. After that, the students will work in groups on jointly identified experiments applying specific solutions to given problems and finally prepare a presentation and write a report about their findings concerning the experiments.

There will be an introductory phase to present basic concepts for the theme, including the necessary foundations.

Lectures will happen through Zoom from our seminar room. The students interested can also join face-to-face in the seminar room.

Leistungserfassung

We will grade the group's reports (80%) and presentations (20%). Note that the report includes documenting the experiments and the obtained results. Therefore, the grading of the report includes the experiments. During the project phase, we will require participation in meetings and other groups' presentations in the form of questions and feedback to their peers.

Termine

The first lecture will take place on October 24, 2023 (Tuesday) in room A-2.8 and remotely via Zoom link*

We will follow the recurrent schedule of:

• Tuesday 13:30 - 15:00, Room A-2.8

* Please email he.xu [at] hpi.de 

Zurück