Automated Closed-loop Tests for Lighting Assistance Systems

Abstract

The options available for testing lighting assistance systems give rise to several challenges. These are addressed in the following. Lacking reproducibility is one of the main challenges in testing lighting assistance systems on a statistical basis as well as on reference drives. Ambient lighting conditions are determined by weather (degree of cloud cover), position of the moon, time of day and season. To perform testing, it would in principle be possible to wait for a specific ambient lighting situation to occur. This possibility, however, is nothing more than a theoretical option because the testing periods are usually set and can only be moved within a narrow time slot in the overall framework. In terms of balancing system performance, overall system testing is shown to be imperative because only by integration in the overall system is it possible to align the calibration parameters in a way that provides the maximum possible level of performance while also creating the brand image that the vehicle manufacturers want to achieve. In the worst-case scenario, the different vehicle derivatives from the same model range result in perceptible differences in the way lighting assistance systems perform because one derivative has been validated in summer and one in winter or one at full moon and another at new moon. The approach so far has been to use a broad statistical data base and a good measure of expertise. This has involved high development costs and latent project risks in every development project.