Augmented integrity in ADAS as edge devices with safety assurance 

Edge devices require flexible management of mixed-criticality applications such as the case of advanced driver-assistance systems (ADAS) with guarantees of real-time safety and cybersecurity. 

 AI bricks as hardware accelerators on the Edge while avoiding compromising the guarantees needed for critical applications. 



  • Low-level Hypervisor supporting AI-based hardware accelerators and flexible management of resources for mixed-criticality systems. 
  • Extension of a low-level hypervisor made by Siemens that has proven to be working and is distributed under GPL. 
  • Ensure safety by runtime monitoring, where monitors are correctly synthesized by the high-level domain-specific language.
  • Domain-specific language supporting design, integration and customization of mixed-criticality applications. 



  • Domain-specific language (DSL) 
  • Hypervisor 
  • Monitors 
  • Verification toolchain 
  • Encryption 




  • High-integrity software and economic integration 
  • Avoid issues related to deployment on numerous different hardware platforms 
  • Economic AI-based bricks to perform high performance tasks for the next generation ADAS/AD. 
  • Provides increased level of safety to ADAS/AD 
  • Support of FPGA-based accelerators