Abstract
Keywords: test automation, infotainment function test, optical inspection, OCR, test infrastructure
Innovative products and new technology are becoming increasingly popular in all areas of life. Due to this fact, many new features are appearing in automobiles every year. In particular, new infotainment components are meeting customers' demands for more entertainment in cars and are making driving more comfortable. The automotive industry is meeting the demands of customers by continuously transferring new multimedia technology into cars and by developing an increasing number of driver support and assistance systems.
As a consequence of developing technology, the quantity and complexity of electronic components in automobiles is increasing enormously. This is resulting in complex test equipment and processes for testing. In this paper, we will present a new testing tool which offers opportunities for testing complex, networked ECUs during early prototype phases, focusing on the automation of tests. Particular focus is placed on verifying optical responses on displays and instrumentation.
The innovative system which is presented provides a full automatic module for monitoring and testing instrumentation without any human intervention. Tests can be set to run over several days, repeating the same test cases several times under exactly the same conditions. The system consists of a camera system, a robot arm, special image processing software and residual bus simulation in a closed box. It is highly-qualified in detecting sporadic faults.
The system offers repeatability and reproducibility plus the advantage of high accuracy, objectivity and a significant number of repetitions over a 24 hour period. This fact gives consideration to faults at an earlier point in time and enables costs to be reduced via the elimination of faults before the start of production.
A high number of test cases applied to ECUs increases the probability of locating faults. Undertaking high numbers of tests manually is not possible due to a lack of resources, lack of time and the fact that there is no guarantee that the tester will be sufficiently objective and precise during every repetition. Automation of the testing processes is therefore unavoidable.
The system which is presented here is supplemented by a new and powerful text recognition algorithm which has been developed by VOLKSWAGEN AG. It is capable of finding text on displays and works without any tedious teach-in steps. Testing is therefore independent of design changes involving the font, character size or background. This additional feature enhances the extremely versatile system and extends the feature list to include needle detection, dynamic needle analysis, symbol recognition, sound analysis and many other options.