The SaS Seminars are a permanent series of open seminars of the Division of Software and Systems (SaS) at the Department of Computer and Information Science (IDA), Linköping University. The objective of the seminars is to present outstanding research and ideas/problems relevant for SaS present and future activities. In particular, seminars cover the SaS research areas software engineering, programming models and environments, software and system modeling and simulation, system software, embedded SW/HW systems, computer systems engineering, parallel and distributed computing, realtime systems, system dependability, and software and system verification and testing.

Two kinds of seminars are planned:

• talks by invited speakers not affiliated with SaS,

• seminars by SaS researchers presenting lab research to whole SaS (and other interested colleagues).

The speakers are expected to give a broad perspective of the presented research, adressing the audience with a general computer science background but possibly with no specific knowledge in the domain of the presented research. The normal length of a presentation is 60 minutes, including discussion.

The SaS seminars are coordinated by Christoph Kessler.

SaS seminars 2025

• 9/6/2025, 13:15 (Turing) - S. Litzinger: Challenges and Solutions in Energy-Efficient Scheduling of Parallel Streaming Applications
• 30/6/2025, 11:00 (Turing) - X. Zheng: "From Probabilistic Testing to Certifiable AI: Large Language Models and Neuro-Symbolic Reasoning for Verifiable Autonomous Systems"

From Probabilistic Testing to Certifiable AI: Large Language Models and Neuro-Symbolic Reasoning for Verifiable Autonomous Systems

Prof. Xi Zheng, Macquarie University, Australia

Monday, 30 June 2025, 11:00 (sharp), room Alan Turing, IDA

Abstract:
Learning-enabled Cyber-Physical Systems (LE-CPS), such as autonomous vehicles and drones, face critical safety and reliability challenges due to the stochastic nature of deep neural networks. In our FSE'22 and TSE'23 studies, we conducted an in-depth investigation into industry testing practices, uncovering significant gaps between current testing techniques and the needs for regulatory assurance. To address these, we introduced two pioneering approaches: regression testing tailored to learning components and scenariobased test reduction for multi-module autonomous systems. This talk will then present our recent progress in bridging probabilistic testing with formal reasoning. I will first cover our FSE'24 work on reviving model-based testing with Large Language Models (LLMs), followed by TSE'24 and ICSE'25 efforts on LLM-driven scenario generation and online testing for uncrewed drone autolanding. Finally, I will introduce our FSE'25 paper NeuroStrata, which leads a neurosymbolic shift from black-box machine learning to white-box, human-understandable reasoning, aiming to improve interpretability, testability, and certification of AI components in safety-critical CPS.

Speaker's bio:
A/Prof. Xi Zheng earned his Ph.D. in Software Engineering from the University of Texas at Austin in 2015. He is awarded Australian Research Council Future Fellow in 2024. Between 2005 and 2012, he was the Chief Solution Architect for Menulog Australia. Currently, he occupies several leadership roles at Macquarie University, Australia: Director of the Intelligent Systems Research Group (ITSEG.ORG), Director of International Engagement in the School of Computing, Associate Professor and Deputy Program Leader in Software Engineering. His research areas include Cyber-Physical Systems Testing and Verification, Safety Analysis, Distributed Learning, Internet of Things, and the broader spectrum of Software Engineering. A/Prof. Zheng has successfully secured over $2.4 million in competitive funding from the Australian Research Council (1 Future Fellow, 2 Linkages and 1 Discovery) and Data61 (CRP) projects focused on safety analysis, model testing and verification, and the development of trustworthy AI for autonomous vehicles. He has been recognized with several awards, including the Deakin Industry Researcher Award (2016) and the MQ Early Career Researcher Award (Runner-up 2020). His academic contributions include numerous highly cited papers and best conference paper awards. He has served as a Program Committee member for leading Software and System conferences, such as ICSE (2026), FSE (2022, 2024) and PerCom (2017-2025), and as PC chair for IEEE CPSCom-2021 and IEEE Broadnets-2022. Additionally, he has taken on the role of associate editor for ACM Distributed Ledger Technologies and editor for the Springer Journal of Reliable Intelligent Environments. In 2023, A/Prof. Zheng is a visiting professor at both UCLA and UT Austin and co-founder of the international workshop on trustworthy autonomous cyber-physical systems. A/Prof. Zheng is a leading co-organizer for the Shonan Meeting (Seminar No. 235) on "LLM-Guided Synthesis, Verification, and Testing of Learning-Enabled CPS" in March 2026 and the Dagstuhl Seminar (2025-01048) on "Advancing Testability and Verifiability of CPS with Neurosymbolic and Large Language Models" in October 2026.

Challenges and Solutions in Energy-Efficient Scheduling of Parallel Streaming Applications

Dr. Sebastian Litzinger, IDA/LiU

Monday, 9 June 2025, 13:15, room Alan Turing, IDA

Abstract:
Scheduling of parallel applications on multicore processors is a complex optimization problem where specific application requirements such as minimum throughput must often be taken into account. For streaming applications, it may thus pay off to compute a schedule statically for repeated execution. In such a case, one will generally seek to achieve a high solution quality at a low cost. This talk will delineate two approaches which aim at reducing the computational effort compared to those guaranteeing an optimal solution: the heuristic transformation of solutions to a relaxed problem and the systematic design of the search space. Evidently, the negative effect on solution quality should be limited when employing these techniques. A second important problem to be discussed is the runtime adaptation of schedules for applications exhibiting dynamic behavior. Given a (near-)optimal initial solution, modifications to the schedule must accommodate shifts in application behavior, e.g., to preserve feasibility, while avoiding a pronounced decline in solution quality.

Speaker's bio:
Sebastian Litzinger is a postdoctoral researcher with the Programming Environments Laboratory (PELAB) at the Department of Computer and Information Science (IDA), Linköping University, Sweden. He received a PhD in computer science from FernUniversität in Hagen, Germany, in 2023 and holds an MA degree in philosophy from Eberhard Karls Universität Tübingen, Germany. His research focuses on parallel computing, in particular scheduling and resource management techniques to enhance the energy efficiency of parallel systems. Furthermore, he is interested in the application of parallelism and, generally, optimization procedures to problems in cybersecurity. Contact him at sglitzinger.github.io.

Liquid Computing

Prof. Jan Madsen, DTU

Thursday, 6 March 2025, 13:15, room Alan Turing, IDA

Abstract:
A key turning point for the evolution of modern computers was the introduction of the general-purpose computer architecture which allowed the computer to be programmed after hardware fabrication. This allowed hardware to be designed bottom-up, providing a hardware independent programming interface which could then be compiled from a high-level programming language, effectively separating the design of hardware (processor) and software (application). Today, microfluidic Lab-on-Chip are still mainly developed as hardcoded applications using passive operations, even though microfluidic technologies have the potential to scale. In this talk we argue that such scaling will require active components and the ability to abstract basic operations to a level similar to that of classical computation, and present a digital microfluidic-based general-purpose Lab-on-Chip that can be programmed after fabrication. We show that parallel execution of multiple protocols are possible, and that EDA tools for MPSoC can be adapted to solve and optimize protocol execution on digital microfluidic Lab-on-Chip, which we call liquid computing.

Speaker's bio:
Jan Madsen is Full Professor in Computer-Based Systems and Head of Department of Applied Mathematics and Computer Science (DTU Compute). He holds a MSc in Electrical Engineering (1986, DTU) and a PhD in Computer Science (1992, DTU). His research interests are in the intersection between computer science and biotechnology, with a special focus on design, modelling, and construction of microelectronic (MPSoC and IoT), microfluidic (Lab-on-Chip) and microbiological (molecular) computing systems, including the development of design automation tools and design methodologies. In these fields, he has co-authored more than 200 peer-reviewed journal and conference papers, 3 books, 13 book chapters, and 3 patents. He has co-supervised 51 PhDs. He was the General Chair of DATE 2018, NOCS 2012, CODES 2001 and the Technical Program Chair of DATE 2007, CODES+ISSS 2011, CODES 2000. Dr. Madsen is a board member of EDAA, member of the Academy of Technical Sciences and its Council for Technology and Society. He was National ICT Expert for EU Horizon 2020. He is member of IEEE and ACM.

Towards Trustworthy and Factual Large Language Models

Prof. Fredrik Heintz, IDA

Tuesday, 4 March 2025, 13:15, room Ada Lovelace, IDA

Abstract:
Europe has taken a clear stand that we want AI, but we do not want just any AI. We want AI that we can trust. This talk will present ongoing research from the EU project TrustLLM which has the goal of developing more factual and trustworthy large language models. To achieve the ambitious objectives of this project, TrustLLM will tackle the full range of challenges of LLM development, from ensuring sufficient quality and quantity of multilingual training data, to sustainable efficiency and effectiveness of model training, to enhancements and refinements for factual correctness, transparency, and trustworthiness, to a suite of holistic evaluation benchmarks validating the multi-dimensional objectives.

Speaker's bio:
Fredrik Heintz is a Professor of Computer Science at Linköping University, where he leads the Division of Artificial Intelligence and Integrated Computer Systems (AIICS) and the Reasoning and Learning lab (ReaL). His research focus is artificial intelligence especially Trustworthy AI and the intersection between machine reasoning and machine learning. He is director of the Wallenberg AI and Transformative Technologies Education Development Program (WASP-ED), Co-director of the Wallenberg AI, Autonomous Systems and Software Program (WASP), Coordinator of the TrustLLM project, and Vice President for AI Research Adra the AI, Data, and Robotics partnership. Member of the Swedish AI Commission. Fellow of the Royal Swedish Academy of Engineering Sciences (IVA).

Composition and Automated Verification of Stateful Security Protocols in Isabelle/HOL

Associate Prof. Sebastian Mödersheim, DTU, Denmark

Friday, 28 February 2025, 10:15, room Ada Lovelace, IDA

Abstract:
Composition is the question: given that two protocols are secure in isolation, are they also secure together - running on the same network and sharing the same key infrastructure? A particular challenge is if the protocols are stateful, e.g., a server that maintains long-term information in a database; if such a server is involved in more than one protocol, then compositional verification requires an appropriate interface between the protocols in terms of assumptions and guarantees by each protocol concerning the database. One can then verify that each protocol is secure in isolation w.r.t. this interface, and obtain that the composed system is secure. The compositionality theorem is formalized in Isabelle/HOL so one can automatically check the requirements to the protocols and thus turn a proof of security for individual protocols into a proof for the composed protocol. To obtain the proof for the individual protocols automatically, too, we have designed the PSPSP tool that uses abstract interpretation to obtain an Isabelle proof for a large class of stateful protocols. Together one can thus automatically obtain a machine-checked proof for many composed stateful protocol systems.

Speaker's bio:
Sebastian Mödersheim received his PhD in 2006 at ETH Zürich (advisor: David Basin). During 2007-2009 he was postdoc at IBM Zürich. Since 2010 he is Associate Professor at DTU in the Software Systems Engineering group (former Formal Methods, former Language-Based Technology headed by Hanne Riis Nielson). His research interests include: formal methods for security and privacy, specifically modelling; automated verification; and compositional reasoning for security protocols. He developed many verification tools (and corresponding languages), including AVISPA/OFMC, AIF, and PSPSP.

Previous SaS Seminars

• 2024
• 2023
• 2021-22
• 2020
• 2019
• 2018
• 2017
• 2016
• 2015
• 2014
• 2013
• 2012
• 2011
• 2010
• 2009
• 2008
• 2007
• 2006
• 2005
• 2004
• 2003
• 2002
• 2001

Page responsible: Christoph Kessler
Last updated: 2025-05-14