Collaborative virtual validation platform for ADAS/AD at KIT (ValKIT)

  • Contact:

    Qais Hamarneh, Maike Schwammberger

  • Funding:

    KIT Matching Fund

  • Startdate:

    01.01.24

  • Enddate:

    31.12.24

Motivation

Due to the increasing degree of automation of vehicles, the complexity of the interaction between the vehicle and the surrounding systems is increasing in the direction of a system of systems (SoS). The validation of such systems, especially their interactions, represents a major challenge for the development of automated vehicles (AVs). To ensure the safety of all road users, identifying and analyzing critical driving scenarios at design time is of utmost importance to ensure the safety and acceptance of AVs. Virtual validation platforms are critical in addressing this challenge, providing scalability, ease of use and safety for individuals. They enable a comprehensive simulation of a vehicle, taking into account its environment, including vulnerable road users, infrastructure and other road users. The integration of numerous components into the virtual platform is a complicated, cross-functional task that often exceeds the capacity of a single researcher. To streamline this process, our goal is to establish a common and collaborative virtual validation platform at KIT that utilizes existing and openly available components and enables swarm testing, as shown in . This project aims to lay the foundation for a versatile validation platform that can serve as a cornerstone for numerous strategic projects in the future, but also offers the opportunity to further expand its capabilities.


Objectives:

  • To develop a versatile and virtual platform for the development and validation of mobility systems (e.g. AVs) as well as for the creation of load spectra and user scenarios. The platform should be open, modular and available to all members of the center.
  • Enabling swarm tests and promoting collaborative work in the virtual environment.
  • Illustrate the potential of shared and collaborative virtual platforms through two key use cases that utilize swarm testing:
  • Collecting and analyzing conflict scenarios. A state is considered a conflict when an agent, whether human or autonomous, cannot choose an action (e.g. turning at an intersection, accelerating, braking...) without violating existing rules.
  • Creation of critical conflict scenarios through a gamification approach in the shared environment and subsequent automated solution or teleoperation in AR/VR/MR