Ground Truth – Core Element of Data-driven AD Development

摘要

The critical path for introducing autonomous vehicles is not the technology as such but rather the development of a metric enabling the homologation, according to a 2013 assessment. Looking at the history of autonomous driving from its beginnings in the 1980s as part of the Eureka-Prometheus project [4], this assessment from 2013 still holds true. Recording of real-world driving data with a DGT reference system and a data-driven development process are certainly decisive building blocks on the way to the road release of a vehicle with autonomous driving functions. The early alignment of this validation process and the necessary software toolchain with the regulatory and certification authorities are key prerequisites for mastering this interdisciplinary challenge.