Author(s):
J. Wasmund, M. Grochow, S. Reglitz, H. Krumm
dSPACE GmbH, Paderborn, NRW, 33102, Germany
Autonomous systems must be able to perform a task while interpreting their environment and often reacting to unforeseen, complex events. Typically, these systems consist of many subsystems to detect the environment (e.g., lidar, radar), subsystems to interpret the collected data, and actuators to perform the desired action, forming a cyber-physical system of systems.
The computational effort required to simulate such systems often exceeds the available resources of a single computing node. It is therefore necessary to run these simulations in a cloud environment to be able to dynamically scale the required resources. However, in later phases of the development process, it is necessary to couple cyber-physical systems with the real target hardware. This paper describes a new approach on how to realize such a coupling between a hardware-in-the-loop simulator and a software-in-the-loop simulator running in a cloud environment based on ED-247 and Kubernetes.