Closely linking physical road testing and virtual testing is the key to develop automated driving functions efficiently. In order to leverage the full potential, we developed the AAI Scenario Cloning module, a tool which automatically processes sensor measurements from real-world test drives.
Our scenario cloning solution allows the extraction of interesting maneuvers from sensor recordings, which in the first stage are converted into reproducible deterministic simulation test scenarios. At a second stage, these concrete test scenarios can furthermore be generalized and converted into parameterizable templates. Insights from physical test drives or virtual endurance tests therefore can be consistently and iteratively reused for the simulation test automation - ready for re-testing after each update of the automated driving software based on the same standards.
Many companies already are in possession of petabytes of sensor recordings (or "measurement data") from physical road testing (or "field operation tests"). The art is to efficiently extract precious insights from these data lakes for repeatable future use. With the AAI solution, it is possible to trace back the trajectories of each traffic participant within the sensor range of the vehicle under test, resulting in a deterministic description of a specific incident as a virtually reusable scenario.
Quantity will not always guarantee quality. With our tool chain we help to use precise KPIs for discovering and extracting only relevant scenarios for the specific customer test cases. We support novelty comparison and criticality tagging (based on aligned criteria from our customers) in order to cleanly manage the huge amount of test cases collected via Scenario Cloning. Let us know your requirements for a customized solution fitting your company's needs!