Programme.
08:45-09:00
Registration and Welcome
Keynote Talk
09:00-10:00
Automated Vehicle Safety Revisited
Philip Koopman
10:00-10:30
Coffee break
Session 1 (Chair: Zakaria Chihani)_
10:30-10:50
A Flow-based Credibility Metric for Safety-critical Pedestrian Detection
Maria Lyssenko, Christian Heinzemann, Matthias Woehrle, Christoph Gladisch and Rudolph Triebel
(L)
10:50-11:10
Enclosing Prototypical Variational Autoencoder for Explainable Out-of-Distribution Detection
Conrad Orglmeister, Erik Bochinski, Volker Eiselein and Elvira Fleig
(L)
11:10-11:30
Language Models as Reasoners for Out-of-Distribution Detection
Konstantin Kirchheim and Frank Ortmeier
(L)
Session 2 (Chair: Mario Trapp)_
11:30-11:45
Bringing Classifiers into Critical Systems: Are we Barking up the Wrong Tree?
Tommaso Zoppi, Fahad Ahmed Khokhar, Andrea Ceccarelli and Andrea Bondavalli
(S)
11:45-12:00
Dual-Process System: an Architectural Pattern for Assurable Autonomous Robots Inspired
by Dual-Process Theory
Krzysztof Czarnecki
(S)
12:00-12:15
Neuro-Symbolic Artificial Intelligence for Safety Engineering
Laura Carnevali and Marco Lippi
(S)
12:15-12:30
Homeopathic poisoning of RAG systems
Katarzyna Kapusta and Boussad Addad
(S)
12:30-13:30
Lunch break
Session 3 (Chair: Jérémie Guiochet)_
13:30-13:50
Uncertainty-aware evaluation of quantitative ML safety requirements
Simon Burton, Benjamin Herd and João-Vitor Zacchi
(L)
13:50-14:10
Bridging the reality gap: assurable simulations for an ML-based inspection drone flight controller
Philippa Ryan, Mathias Von Essen, Liam Shackley and John McDermid
(L)
14:10-14:30
LLMs in Web Development: Evaluating LLM-Generated PHP Code Unveiling Vulnerabilities and Limitations
Rebeka Tóth, Tamas Bisztray and László Erdődi
(L)
Session 4 (Chair: Philippa Ryan)
14:30-14:50
14:50-15:05
Continuous Development and Safety Assurance Pipeline for ML-based Systems in the Railway Domain
Marc Zeller, Thomas Waschulzik, Carmen Carlan, Marat Serahlazau, Bahlmann Claus, Zhiliang Wu, Sigurd Spieckermann, Denis Krompass, Lucia Diez Robles, Simon Geerkens, Christian Sieberichs, Konstantin Kirchheim and Batu Kaan Özen
Learning-based Safety Monitor for Telltales in Vehicle Instrument Clusters
Cornelius Buerkle, Fabian Oboril and Kay-Ulrich Scholl
(L)
(T)
15:05-15:20
Combining Uncensored and Censored LLMs for Ransomware Generation
Muhammad Mudassar Yamin, Ehtesham Hashmi and Basel Katt
(T)
15:20-15:35
Accident Case Study Analysis of Developmental Automated Driving System (ADS) Collision (Part I and II)
Sanjeev Appicharla
(T)
15:35-16:00
Coffee break
Community Debate (Chair: Orlando Avila-García/Simon Burton)
16:00-16:15
Detecting and Mitigating Errors in Neural Networks
Uwe Becker
(T)
16:15-16:30
Enhancing LLM Security through Ontology-Based Assurance Cases
Tomas Bueno Momcilovic, Dian Balta, Beat Buesser, Giulio Zizzo and Mark Purcell
(T)
16:30-17:00
17:00-17:10
Plenary Discussion: "Safety Engineering, Foundation Models and LLMs - Dr. Jeykill and Mr. Hyde"
WAISE Wrap Up