ASML Internship:Epistemic Reasoning on Knowledge Graphs for Fault Diagnosis

Supervisor: Daira Pinto Prieto (d.pintoprieto@vu.nl)

Background

Fast and efficient diagnosis of ASML equipment failures is a critical business goal. However, diagnosing complex cyber-physical systems is a challenging task. Diagnostic assistants can support service engineers by suggesting actions to identify the root cause of a failure or restore the system to operation. The reasoning process behind such assistants may vary depending on the system, the available information, and the diagnostic goal. Since diagnostic scenarios are highly heterogeneous, a reasoning strategy that is effective in one situation may be ineffective or even misleading in another. For this reason, the investigation of alternative reasoning methods for diagnostic support systems remains an active research topic, and this is the area in which this MSc thesis aims to contribute.

Diagnostic problems typically start from a set of observed symptoms together with a knowledge base describing the system under diagnosis. Knowledge graphs provide a suitable representation for such knowledge, encoding facts as subject-predicate-object triples (for instance using RDF language) and enriching them with ontological concepts and rules through OWL. Classical Description Logic reasoners can then derive implicit knowledge, detect inconsistencies, and integrate observations with domain knowledge. While this type of factual reasoning is essential, diagnosis also requires generating and managing hypotheses under uncertainty. In other words, it must account not only for what is known, but also for what is unknown, plausible, or still under investigation.

This project proposes to explore Dynamic Epistemic Logic (DEL), or a computationally feasible approximation of it, as a complementary reasoning layer on top of a domain-specific knowledge graph. DEL provides formal mechanisms to represent what is known, what is supported by evidence, and what is believed, as well as to model information updates, enabling the generation and revision of diagnostic hypotheses as new evidence becomes available. The goal is to investigate whether epistemic reasoning on top of knowledge graphs can support a richer and more flexible diagnostic process, providing service engineers not only with the assistant’s primary hypothesis, but also with information about directions that have been ruled out, remain plausible but unproven, or are currently unknown due to gaps in the available evidence.

Description

Understand the goals and reasoning process of the ASML diagnostic application that we aim to enrich. Given a domain-specific knowledge graph describing failure modes and the physical and functional aspects of the system, together with representative diagnostic scenarios (available symptoms and the current results produced by the application), investigate which information in the knowledge graph can be leveraged to refine or enrich the current diagnostic results. Develop and evaluate epistemic reasoning algorithms for generating diagnostic hypotheses and other epistemic states, analyze their correctness and scalability, and propose semantic approximations when needed. Such approximations may consist of semantic measures or graph queries that retrieve similar information in a more efficient manner. The domain-specific knowledge graph, diagnostic scenarios, and baseline epistemic logic algorithms will be provided.

Research questions:

The expected outcome is a reasoning pipeline that takes as input the observations and results generated by an industrial diagnostic application, together with a knowledge graph describing the system under diagnosis, and produces a richer epistemic characterization of the diagnostic scenario. This characterization should include what is currently believed, ruled out, insufficiently supported by evidence, or unknown due to gaps in the available information.

Literature

Hogan, A., Blomqvist, E., Cochez, M., D’Amato, C., de Melo, G., Gutierrez, C., Kirrane, S., Labra Gayo, J.E., Navigli, R., Neumaier, S., Ngonga Ngomo, A., Polleres, A., Rashid, S.M., Rula, A., Schmelzeisen, L., Sequeda, J., Staab, S., Zimmermann, A. Knowledge graphs. ACM Computing Surveys 54(4), Article 71 (2022), [PDF].

Baltag, A., Bezhanishvili, N., Özgün, A. et al. Justified belief, knowledge, and the topology of evidence. Synthese 200, 512 (2022), [PDF].

Pinto Prieto, D., Tiddi, I., Schlobach, S. Enabling formal epistemic reasoning in diagnostic systems using semantic technologies. In: Knowledge Engineering and Knowledge Management - 25th International Conference, EKAW 2026, Proceedings. Springer, 2026, [PDF].

Prerequisites