Safety System Safety

Safety architectures enabling autonomous driving

Automated vehicles - both fully automatic and fully automatic or semi-automatic equipped with driver assistance functions - are currently a prioritized topic for the automotive industry. The stated goal is the autonomously operating safe vehicle, which is to relieve the traffic either on roads and the nerves of the user or on rails and safe driving areas the operator. In order to achieve the goal in terms of safety, first of all the human-machine cooperation must be illuminated. For this purpose, an older, modified model from the flight guidance is shown and the effective loops, in which the driver is involved (recognize situation, understand, act accordingly), discussed. Based on the simple model with the effective loops, the supportive effect of current driver assistance functions can already be represented so far that autonomy would be achievable.




However, there are limits to this autonomy, which are set forth in particular by the sensors,
the corresponding diagnoses / monitors and the availability and reliability of the functions
and the diagnoses / monitors. These limits are for the most part already described as
Functional Safety. But even if the functional safety requirements are fully met,
there may still be gaps if the response time of the operator / driver has to be resorted
to in the event of a breakdown and this has already been very restricted by the situation
or architecture of the vehicle

Assurance Veritas Quality (AVQ)

AVQ GmbH is an independent consulting company specialized in functional and system safety of complex technical systems. Our clients value both, our expertise and our pragmatic approach. Our long-term customer relationships are characterized by mutual trust and respect. This is how we create a framework where everyone can realize their potential. This allows us to learn from each other and grow together.

Professionalism, integrity, humanity and passion are the base of our daily trading. These values enable us to help our customers meeting their challenges and strengthen their competitiveness.

AVQ Safetymanagement


  • ISO26262 Tool qualification of 3D measuring systems (optical, radar, laser) and driving simulation systems
  • Functional Safety Concept and Evidence of Artificial Intelligence AI
    • Cognitive systems in general
    • Expert systems - e.g. on FPGA
    • Machine learning to deep learning of neural networks
  • Qualification of safety-related data and acquisition / measurement eg precise 3D map of transport infrastructure vs. Safety-relevant driving information
  • Safety of use (SOTIF) of driver assistance systems SAE J3016 Level 2-3 to highly automated driving SAE J3016 Level 4
  • Modular architectures for fail operational requirements fully automated driving VAF
  • Functional safety of big data sensor fusions and actuator control systems in highly automated HAF driving
  • System Safety for autonomous off-road driving
  • Safety limits of sensors and definition Defensive driving "Functional Accuracy" of Actuator Control Systems in Commercial Vehicles
  • Side effects of medicine 4.0