Pfaffenhofen an der Roth, 6 March 2025. At this year's embedded world from 11 to 13 March, NewTec GmbH is presenting its new AI services. In addition to an AI Safety Check, the offering includes extensive consulting services for manufacturers and developers of embedded systems.

When AI technologies are to be integrated into embedded systems in which a failure of the technology entails risks, manufacturers face particular challenges. Unlike conventional deterministic systems, it is not possible to determine unambiguously whether AI technologies meet functional safety requirements. The performance of AI technologies is determined by a number of random variables – with corresponding variance. Statements about their performance and accuracy can therefore only be made in the form of statistical estimates. Nevertheless, for high-risk systems with safety-relevant AI products, such as lifts, vehicles or critical infrastructure, the EU AI Act requires "an appropriate level of accuracy".

With an AI Safety Check, NewTec now supports manufacturers of embedded systems in identifying potential critical AI failures and determining the appropriate level of accuracy and the performance required to prevent failures. The accuracy of the AI technology used (or planned) is then determined using statistical methods.

This procedure, supported by TÜV Rheinland, gives manufacturers clarity as to whether the AI technologies used meet functional safety requirements and regulatory provisions.

The AI Safety Check is a central component of NewTec's AI services, which comprise:

  • Identification of specific failure modes of the AI technology used
  • Analysis of the possible error propagation of the AI technology along the information flow in the system context
  • Analysis of the possible risks arising from failures that may be caused by the AI technology
  • Derivation of target values for the functional suitability of the AI technology (e.g. required accuracy)
  • Validation of the target values with regard to their achievability in the system context with the necessary confidence
  • Hardware analysis to determine the probability of random faults on complex hardware used to execute AI technologies
  • Analysis of the specific development processes of the AI technology used with regard to ensuring systematic capability
  • Preparation of the results and of a presentation to assessment bodies

Further information on the AI services is available at the NewTec stand in Hall 5, stand number 5-170.