概述

Current data logging approaches in ADAS/AD development are not sustainable. The need for data is the main driver of development and validation costs, which surpass billions of dollars globally on the way to autonomy. But only a limited number of logged traffic situations can be used in AD development. Will the barrier posed by data decrease in the years to come? Or will we find a way around it? Is the right path higher efficiency or a radical technology breakthrough? Most likely, it will be a combination of both: In the short term, higher efficiency will help decrease costs. And we can already see the first signs for alternatives to the "machine learning beast" demanding all the data.

This recording will provide insights into how the role of data logging will evolve in coming years and what impact it will have on data-driven development and vice versa. The recording will present the main challenges, both technical and non-technical, that are setting the course for data-driven development. There is a dynamic development in corresponding technologies that will drive final costs down and quality up, which promises improvements and more viable business cases.