Requirement-driven testing of driverless transport systems
The increasing complexity of driverless transport systems due to the introduction of AI applications such as people detectors, data-driven planning and control procedures poses a challenge for the verification and validation (V&V) methods used in the automation industry. For example, DIN EN ISO 3691-4, which is applied in the area of driverless industrial trucks, does not address the use of AI applications in any way.
The focus is on mechanical contact detection systems, short braking distances and thus low speeds. AI applications can detect obstacles and people at an early stage and thus enable higher speeds and throughput rates if safety can be guaranteed. This requires new standards, which are currently being developed in committees as part of the German standardization roadmap for artificial intelligence. Furthermore, standards for verification and validation must also be created, which support the use of artificial intelligence for validation.
Robo-Test is part of the standardization committee “Roadmap Artificial Intelligence” and contributes 25 years of experience in the industrial environment and in research for reliable and safe automation and AI to the German standard for Artificial Intelligence. In addition, Robo-Test contributes to VDI 4004 Part 1 and Part 2, which address test processes, their planning and digital documentation, thus laying the foundation for AI-based automation of assurance.
Read the our Whitepaper out now in german: Download