Safety-related risk assessment of a cyber-physical model system for industry 4.0 applications

  • Project number: F 2497
  • Institution: Federal Institute for Occupational Safety and Health (BAuA) / Universität Stuttgart, Institut für Automatisierungstechnik und Softwaresysteme
  • Status: Completed Project

Description:

Modern production systems exhibit distinctive characteristics such as high complexity, heterogeneity of individual components, autonomy, and reconfigurability. This project addressed key challenges that safety engineers will increasingly face in the near future when performing risk assessments under these conditions.

The main focus of the project was to analyse the extent to which existing risk assessment methods are applicable to Cyber-Physical (Production) Systems (CPPS). The methods were classified, and their strengths and limitations were analysed, along with their potential for integration in the context of CPPS. To validate the findings, a hardware/software demonstrator was developed that simulates a smart factory environment and illustrates advanced risk assessment techniques for industrial CPPS. The demonstrator highlights the limitations of traditional methods in capturing dynamic risks and emphasises the benefits of advanced techniques such as probabilistic model checking and intelligent fault injection. By simulating various configurations, it provides insights into trade-offs between efficiency, safety, and flexibility, serving as both an educational and research tool for the further development of risk assessment methodologies in smart manufacturing environments.

The results underscore the necessity of complementing traditional approaches with more dynamic techniques, including probabilistic model checking, AI-assisted strategies, digital twins, and intelligent fault injection. Based on these findings, the project developed an initial roadmap to support further research and the advancement of risk assessment methods for CPPS.

Publications

A practical approach to evaluating the adversarial distance for machine learning classifiers

Publishing year: 2024

Suchergebnis_Format Essay

Find out more

Investigating the Corruption Robustness of Image Classifiers with Random p-norm Corruptions

Publishing year: 2024

Suchergebnis_Format Essay

Find out more

Utilizing Class Separation Distance for the Evaluation of Corruption Robustness of Machine Learning Classifiers

Publishing year: 2022

Suchergebnis_Format Essay

Find out more

An overview of the research landscape in the field of safe machine learning

Publishing year: 2021

Suchergebnis_Format Essay

Find out more

Aspects of industrial CPS critical for risk assessment methods

Publishing year: 2021

Suchergebnis_Format Essay

Find out more

Krakenbox: deep learning-based error detector for industrial cyber-physical systems

Publishing year: 2021

Suchergebnis_Format Essay

Find out more

Safe machines with - or despite - artificial intelligence

Publishing year: 2021

Suchergebnis_Format Essay

Find out more

Deutsche Normungsroadmap. Künstliche Intelligenz

Publishing year: 2020

Suchergebnis_Format Cooperation

Find out more

German Standardization Roadmap Artificial Intelligence

Publishing year: 2020

Suchergebnis_Format Cooperation

Find out more

Industry 4.0: Emerging challenges for dependability analysis

Publishing year: 2019

Suchergebnis_Format Essay

Find out more

Sicherheitsnachweis bei digital vernetzten Maschinen und Anlagen in wandelbaren Fabriken

Publishing year: 2019

Suchergebnis_Format Essay

Find out more

Further Information

Contact

Unit 2.6 "Workplaces, Safety of Machinery, Operational Safety"

Phone: +49 231 9071-1971
Fax: +49 231 9071-2070