Events and Trade Shows

WS#5 AI and Data-Driven Methods for Robotcs Safety

Moderators: Dr. Magnus Albert, SICK AG, Germany, Dr. Björn Matthias, ABB, Germany, Dr. José Saenz, Fraunhofer IFF, Germany

Main questions to be answered

This workshop delves into the transformative role of Artificial Intelligence in enhancing safety within robotic systems, with a focus on the role of data‐driven approaches. As AI technologies become increasingly embedded in autonomous and collaborative robots, ensuring their safe and reliable operation demands new paradigms in safety engineering. This session explores how data‐driven approaches and machine learning techniques can be systematically integrated into robotics systems and enhancing performance and efficiency while maintaining an appropriate level of safety. The main questions to be answered are:

• Which established safety methodologies are applicable to AI and data‐driven models within robotics systems?
• Which particular role do data‐driven approaches have? What changes for the safety assurance for datadriven models?
• Which regulatory frameworks are already in place and which ones are needed in the future?

WS Content

This workshop will open with an introduction, followed by three expert presentations that explore diverse facets of AI and data‐driven methods in robotics safety. Three distinguished speakers will share insights from industry, regulatory, and research perspectives, addressing both the design‐time and runtime integration of AI into robotic systems and the safety challenges which come along. A particular emphasis will be placed on autonomous robots, where the complexity and unpredictability of environments demand robust, adaptive safety mechanisms.

The presentations will lay the groundwork for a moderated discussion between the speakers and experts from the audience. Key topics will include the requirements for high‐quality datasets, techniques for evaluating data reliability, and the lifecycle management of data‐driven models used in safety applications. The session will also delve into the Verification and Validation (V&V) of AI‐based safety systems. Participants will engage with forward‐looking discussions on how data‐driven approaches are reshaping the safety landscape in robotics, including the potential for new standards and certification pathways. This workshop offers a valuable opportunity for professionals to deepen their understanding, exchange ideas, and contribute to the evolution of safe and intelligent robotic systems.

WS Organisation

Three focus talks will open the session and pave the ground for an interactive discussion of the speakers and the audience. The workshop will utilize the so‐called fish‐bowl discussion format to actively engage experts from the audience, while making sure that we keep focus.

Intended outcome

Gain experience from different fields to be applied in robotics.

• Strengthen the euRobotics network and engage members especially of TG safety, TG industrial robotics, TG Standardisation and TG Laboratory Robotics.
• Potential outcome could be a whitepaper on challenges and particularities of data‐driven approaches within the safety of robotics systems.

Speakers

• Robert Puto, Senior Vice President TÜV Süd Hong‐Kong
• Alexander Hirsch, Senior AI Safety Engineer, NVIDIA
• Johan Stärner, R&D Principal Engineer, ABB Robotics

Topic Groups and/or Innovation Networks involved

TG safety, TG industrial robotics, TG Standardisation and TG Laboratory Robotics