Live Migration of a 3D Flash LiDAR System between two Independent Data Processing Systems with Redundant Design

Self-driving and self-flying vehicles have the ability to drive respectively fly independently without the intervention of an operator. For this purpose, these vehicles need sensors for environment perception and data processing systems, which are safety-critical, to process the obtained raw data from these sensors. However, if such safety-critical systems fail, this can have fatal consequences and can affect human lives and/or the environment, especially in the case of highly automated vehicles. A total failure of these systems is one of the worst scenarios in an automated vehicle. Therefore, such safety-critical systems are often designed redundantly in order to prevent a total failure of environment perception. In order to ensure that the operation of the vehicle can continue safely, however, the live migration from one system to the other must be carried out with as little downtime as possible. In our publication, we present a concept for a 3D Flash LiDAR live migration between two independent data processing systems with redundant design. This concept provides a solution for highly automated vehicles to remain fail-operational in case one of the redundant data processing systems fails. The results obtained from the implemented concept, without specifically addressing performance, are also provided to demonstrate feasibility.