Automated Statistical Validation using Big Data

Authors

M. Pavelka, M. Jonis, T. Kletecka, P. Krejci, J. Olšina, M. Zima, Valeo

Summary

The statistical validation of Advanced driver-assistance systems (ADAS) and Autonomous Driving (AD) systems is data demanding where data processing and state monitoring are infeasible to be done manually as the amount of recorded data acquired for the purpose of a validation is on a big data petabytes scale. A robust end-to-end automated process is presented that includes data collection campaign planning, data recording, labeling, logistics, and data monitoring. Advanced mechanisms of toolchain coordination and data monitoring are mandatory to have an instant clear view of the campaign status to control the campaign in an efficient way. The benefits of the proposed solution are cost reduction, human error risk reduction, scalability and modularity allowing fast customization to project needs.