The purpose of electronic control units (ECUs) with optical displays, such as instrument clusters in vehicles with a speedometer, tachometer, etc., is to interact with the user. Whereas ECUs with purely electronic or mechanic interfaces can be tested easily via the standard physical interfaces of an HIL system, instrument clusters require an additional optical interface to capture the changes on the displays. Because of this, verification of the instrument clusters is often still done “manually” by developers watching the displays. This does not guarantee sufficient precision, though, because the human eye is prone to errors and can tire out. In addition, such tests cost a lot of time and money. It is also not possible to reproduce these tests precisely or to automate them. This is why dSPACE has linked camera-based real-time image recognition with its hardware-in-the-loop simulation technology. This combination allows single tests of individual ECUs with visual components, as well as tests of these ECUs together with the entire vehicle electronics. The advantages of automated and trackable testing of optical displays in real-time lead to better quality and quantity of the tests and, ultimately, of the ECUs themselves.
The first pilot project was developed in the USA by dSPACE Inc. together with General Motors. The configuration is now being used successfully by many other customers – for single ECU tests as well as for integration tests.
When testing ECUs with optical displays, the challenge lies in coupling the camera used for real-time image capturing with a fast computing unit. The test system calculates the necessary input signals for the ECU, controls the ECU via discrete interfaces or bus interfaces, and prepares a feedback channel for the camera. With automatic test sequences, the visual signal on the HIL system that was interpreted by the camera is compared to the input signals. dSPACE simulators have the necessary I/O channels and bus channels to operate the ECUs, including the feedback channel to connect the camera to the simulator. With the help of high-performance processor boards, the environment (incl. the vehicle model and driver model) and the I/O of the ECU are simulated in real time.