Presentation at MSCPES 2021 “Explainable Cyber-Physical Energy Systems based on Knowledge Graph”
On Tuesday, 18 May 2021, one paper from the SemSys group was presented in the 9th workshop on Modeling and Simulation on Cyber-Physical Energy Systems. The paper results from cooperation with Siemens and is entitled: “Explainable Cyber-Physical Energy Systems based on Knowledge Graph.” The authors are: Peb R. Aryan, Fajar J. Ekaputra, Marta Sabou, Daniel Hauer, Ralf Mosshammer, Alfred Einfalt, Tomasz Miksa, Andreas Rauber.
The paper is the continuation of our previous paper on the same venue. In this paper, we presented the design and implementation result of the ExpCPS framework. The framework consists of ExpCPS ontology and algorithms to infer causality relationships and generating explanations from these causality relations.
Although the time slot for the presentation is only around 10 minutes, it successfully garners exciting questions from the audience:
- Can it explain two events that happen at the same time, as they might cause each other?
- Are temporal delays taken into account?
- How to differentiate between failures that happen due to physics or cyber events?
The last question is easier to answer as it is where semantics is good at given sufficient knowledge captured. The first two questions are part of ongoing work on formal analysis of the causality representation.
The workshop was organized as an online event from 3 PM to 8 PM and divided into three sessions. The first session was about models and simulations, where we presented this paper. The second session was mainly about the security aspect of the CPES. Finally, the last session was about energy domains aside from electricity, such as hydraulics and gas.
Keyword: explainability, smart grid, knowledge graph