Objective & context
The validation of functions which are not highly automated (SAE level 0-2) is done based on mature processes and methods e.g. simulations, driving tests under controlled frameworks and with small vehicle fleets.
For the validation of highly automated functions (from SAE level 3), there is currently no standardized release process as a wide spread of measuring data under real driving conditions is needed. The validation of highly automated functions is thus economically and methodically not feasible with the present measures for validation.
Shadow Mode is one main building block that is going to be used for the validation of highly automated driving functions. Shadowing is generally considered as the analysis of functions that are implemented in the car and simulate driving actions/decisions, but do not execute them. The software (components) functions are monitored in their behavior and relevant data is stored and transferred to a backend if abnormalities occur (complex rating possibility). The aggregation of this data enables a fast evaluation of errors and anomalies under real driving conditions and enables a fast validation of the functions.
- Decrease of costs in the development of functions and in the validation of functions
- Usage of data in (re-) simulation to reduce real driving tests
- Overall reduction of costs and complexity in the development of functions