Automated Tests for automotive Instrument Panel cluster Based on Machine Vision

This paper approaches the use of machine vision as an automation tool for verification tests in automotive Instrument Panel Cluster (IPC). A computer integrated with PXI modular instruments, machine vision software and Integrated Development Environment (IDE) composes the test system. The IPC is verified in closed-loop using the Hardware-in-the-Loop (HiL) technique in which the HiL system simulates all Electronic Control Units (ECUs) that interact with the IPC. Every simulated ECUs signals are sent to the IPC over CAN (Controller Area Network) bus or hardwired I/O using PXI modules integrated with IDE and its responses are captured by cameras. Using machine vision such images are subjected to Digital Image Processing (DIP) techniques as pattern matching, edge detection and Optical Character Recognition (OCR), which can be applied to interpret speedometer, tachometer, fuel gauges, display and warning lights. The results obtained after data processing are compared to the expected ones, which are defined by the IPC technical specifications. The use of machine vision makes possible the automation of clusters verification test bringing advantages such as increased reliability, repeatability, cost and testing time reduction.