Keynotes

Keynote Web of Things (WoT)

Speaker: Dominique Guinard (EVRYTHNG)

Title: The Web of Things from Research to the Real-World

Abstract: In this talk Dom Guinard will provide a bit of history of how and why the Web of Things all started. He will then explain how EVRYTHNG and other companies moved the concepts to reality, teaching billions of Things to speak Web. We'll focus on applications, challenges and the way forward.

Keynote Knowledge Graphs (KG)

Speaker: Armin Haller (Australian National University)

Title: Knowledge Graph Engineering

Abstract: In this keynote Armin Haller will discuss industry and academic best practices of creating knowledge graphs. He will discuss some of the challenges that designers and users are facing when trying to build these graphs from scratch and/or by using existing data sources. Armin will also present some results from an analysis of the openly available knowledge graphs on the Web, with a particular focus on how interlinked their data and schemas are. He will conclude with some of the experiences gained and challenges faced in a two-year project to build a Knowledge Graph for the Australian Government and discuss some of the open-source tools that his group developed in the process.

Keynote Multi-Agent Systems (MAS)

Speaker: Birgit Vogel-Heuser (Technical University Munich, Institute of Automation and Information Systems)

Title: How to use Knowledge and Learning in Multi Agent Systems for dependable Field Level Control to realize Industry 4.0

Abstract: Industry 4.0 systems require intelligent behavior by machines themselves in the best case in an automatic up to autonomous way. It shall realize at least two senarios namely Adaptive Factory (AF) and Order controlled production (OCP). The talk will introduce both a model based approch to include engineering knowledge as well as a data driven approach that acquire knowledge during operation. Both cases need to fulfil dependability and real time constraints. How to meet those will be introduced using successful pattern.

Keynote Trustworthy AI

Speaker: Sonja Zillner (Siemens AG)

Title: Trustworthiness of Industrial Artificial Intelligence

Abstract: Trustworthiness of Industrial Artificial Intelligence is receiving increasing awareness among the European AI community.  This is driven on the one side by the draft proposal of the European Commission AI regulation. On the other side, customer’s trust in products and services has always been a high-ranked success criteria for industrial (AI) offerings.  The community especially emphasizes the importance of the issues of transparency and explainability of AI, the system’s reliability and the bias of the systems, usage of confidential data and data privacy of the applications. In particular, critical applications that are considered to contain high-risks requiring conformity assessment of trustworthiness. And to ensure conformity with trustworthy requirements, a wide range of AI technologies are needed to mitigate potential risks.

In this keynote, trustworthy AI requirements in the context of Industrial Artificial Intelligence will be introduced. This will be complemented by examples of how AI technologies, such as Explainable AI, Robust / Safe AI, Data Privacy, etc., are used to ensure conformity with trustworthy requirements as well as by an outlook into research and innovation challenges in the area of trustworthy and industrial AI.