Proposal of Automated Inspection Using Camera in Process of VIN Validation

Abstract

After recording the Vehicle Identification Number (VIN) on the chassis, a procedure of validation is indispensable, to ensure that the code be correctly recorded. The automotive sector utilizes this constantly, trying to eliminate mistakes, once that the VIN is utilized in world scale and these errors bring many troubles to the consumers, and consequently for the brand. In some cases, it is not available, the minimum requirements of reliability necessary for the inspection process. In cases where the inspection is performed manually, with an operator making a visual conferencing, there is risk of commercialize a vehicle with a different VIN of the one contained in the documents, owing a human mistake. This paper proposes the automation of the inspection process, using computational vision in data validation transcribed to the chassis, and low-cost components to read the VIN recorded in the chassis, comparing it to the previously authorized code, to increasing the quality control and avoiding future problems. We subject the system to experimental tests and find efficiencies in the inspection recording VIN. We increased reliability in the process, including a pre-validation of the code to be recorded, where the machine is not authorized to perform recording without requirements validated.