Monika Seisenberger
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Monika Seisenberger has a total of 37 co-authors. Here are the top co-authors by number of shared publications:
- Xiuyi Fan
Nanyang Technological University
6 shared publications - Jamie Duell
Nanyang Technological University
3 shared publications - Ulrich Berger
Swansea University Swansea UK
3 shared publications Andrew Lawrence
Siemens Mobility Chippenham
2 shared publications- Andrew Lawrence
Siemens Mobility, Research and Development
2 shared publications - Fabio Caraffini
Swansea University
2 shared publications
Latest Publications
- Open Access
- 16 October 2026
- Published by Springer Nature
- p. 130-149
Formal Assurance for Railway Signalling: Expanding SAT Proof Checking to Certify Z3’s Inequality-Based SMT Proofs for ERTMS
Safety-critical railway systems such as Railway interlockings and the European Rail Traffic Management System (ERTMS) require strong, independently checkable guarantees of correctness. This requirement is reinforced by the new standard governing the use of software in the railway domain, EN50716 [18], which firmly promotes formal verification as part of an end-to-end assurance process. Many modern railway verification workflows rely on SMT solvers, such as Z3, to discharge safety properties expressed in propositional logic or its suitable extensions. In this paper, we show how a Z3-based verification toolchain can be extended with proof checking via independent validation of the solver’s output, thereby increasing trust in the overall verification process. To this end, we present a bespoke proof-checking framework for railway verification, capable of handling propositional logic, as well as inequality reasoning based on the Farkas lemma. The latter plays a major role in ERTMS logic, where positioning constraints are often expressed as linear inequalities. We demonstrate the application of the Farkas lemma through a small case study, showcasing the effectiveness of the tool. The methodology developed for handling Farkas certificates can serve as a foundation for incorporating additional SMT reasoning rules. The proof-checking framework has been formally specified in Rocq, with the checker extracted from the formalisation. This approach has been developed together with our industrial partner. It is concise, tailored to their framework, and independent from other larger proof checking mechanisms.
- 16 October 2026
- Published by Springer Nature
- p. 130-149
Open AccessFormal Assurance for Railway Signalling: Expanding SAT Proof Checking to Certify Z3’s Inequality-Based SMT Proofs for ERTMS
Safety-critical railway systems such as Railway interlockings and the European Rail Traffic Management System (ERTMS) require strong, independently checkable guarantees of correctness. This requirement is reinforced by the new standard governing the use of software in the railway domain, EN50716 [18], which firmly promotes formal verification as part of an end-to-end assurance process. Many modern railway verification workflows rely on SMT solvers, such as Z3, to discharge safety properties expressed in propositional logic or its suitable extensions. In this paper, we show how a Z3-based verification toolchain can be extended with proof checking via independent validation of the solver’s output, thereby increasing trust in the overall verification process. To this end, we present a bespoke proof-checking framework for railway verification, capable of handling propositional logic, as well as inequality reasoning based on the Farkas lemma. The latter plays a major role in ERTMS logic, where positioning constraints are often expressed as linear inequalities. We demonstrate the application of the Farkas lemma through a small case study, showcasing the effectiveness of the tool. The methodology developed for handling Farkas certificates can serve as a foundation for incorporating additional SMT reasoning rules. The proof-checking framework has been formally specified in Rocq, with the checker extracted from the formalisation. This approach has been developed together with our industrial partner. It is concise, tailored to their framework, and independent from other larger proof checking mechanisms.
- 16 November 2025
- Published by Springer Nature
- p. 78-87
Creating Synthetic Test Data for Rail Design Tools – The Case of Linear Scheme Plans
- 16 November 2025
- Published by Springer Nature
- p. 78-87
Creating Synthetic Test Data for Rail Design Tools – The Case of Linear Scheme Plans
- 17 April 2025
- Published by Springer Nature
- p. 320-333
A Multi-Agent System for Optimal Train Scheduling in Single-Track Railways
- 17 April 2025
- Published by Springer Nature
- p. 320-333
A Multi-Agent System for Optimal Train Scheduling in Single-Track Railways
- 21 April 2024
- Published by Springer Nature
- p. 413-424
A Comparison of Global Explanations Given on Electronic Health Records
- 21 April 2024
- Published by Springer Nature
- p. 413-424
A Comparison of Global Explanations Given on Electronic Health Records
- 27 September 2023
- Published by Springer Nature
- p. 211-224
A Tool-Chain for the Verification of Geographic Scheme Data
- 27 September 2023
- Published by Springer Nature
- p. 211-224
A Tool-Chain for the Verification of Geographic Scheme Data
- Open Access
- 4 September 2023
- Published by Elsevier in Finance Research Letters
- Vol. 58
Fund performance evaluation with explainable artificial intelligence
- 4 September 2023
- Published by Elsevier in Finance Research Letters
- Vol. 58
Open AccessFund performance evaluation with explainable artificial intelligence
- 5 June 2023
- Published by Springer Nature
- p. 120-124
Batch Integrated Gradients: Explanations for Temporal Electronic Health Records
- 5 June 2023
- Published by Springer Nature
- p. 120-124
Batch Integrated Gradients: Explanations for Temporal Electronic Health Records
- 17 October 2022
- Published by Springer Nature
- p. 246-268
Safe and Secure Future AI-Driven Railway Technologies: Challenges for Formal Methods in Railway
- 17 October 2022
- Published by Springer Nature
- p. 246-268
Safe and Secure Future AI-Driven Railway Technologies: Challenges for Formal Methods in Railway
- 26 September 2022
- Published by Springer Nature
- p. 411-420
Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers
- 26 September 2022
- Published by Springer Nature
- p. 411-420
Towards Polynomial Adaptive Local Explanations for Healthcare Classifiers
- Open Access
Towards a Shapley Value Graph Framework for Medical peer-influence
eXplainable Artificial Intelligence (XAI) is a sub-field of Artificial Intelligence (AI) that is at the forefront of AI research. In XAI, feature attribution methods produce explanations in the form of feature importance. People often use feature importance as guidance for intervention. However, a limitation of existing feature attribution methods is that there is a lack of explanation towards the consequence of intervention. In other words, although contribution towards a certain prediction is highlighted by feature attribution methods, the relation between features and the consequence of intervention is not studied. The aim of this paper is to introduce a new framework, called a peer influence framework to look deeper into explanations using graph representation for feature-to-feature interactions to improve the interpretability of black-box Machine Learning models and inform intervention.
Open AccessTowards a Shapley Value Graph Framework for Medical peer-influence
eXplainable Artificial Intelligence (XAI) is a sub-field of Artificial Intelligence (AI) that is at the forefront of AI research. In XAI, feature attribution methods produce explanations in the form of feature importance. People often use feature importance as guidance for intervention. However, a limitation of existing feature attribution methods is that there is a lack of explanation towards the consequence of intervention. In other words, although contribution towards a certain prediction is highlighted by feature attribution methods, the relation between features and the consequence of intervention is not studied. The aim of this paper is to introduce a new framework, called a peer influence framework to look deeper into explanations using graph representation for feature-to-feature interactions to improve the interpretability of black-box Machine Learning models and inform intervention.