Development of an Algorithm to Automate Fault Analysis in the Automotive Industry

Industry tries to control all the variables involved in its processes in order to prevent its most varied products from deviating from the conformity normally recommended by the quality teams. When a defect occurs in a given product, time, manpower and, above all, money are often wasted, because a production line that stops is much more expensive than one that runs without defects. The aim of this research is to develop an algorithm capable of automating the analysis of electrical faults in a vehicle. To this end, the results of a survey of the faults repaired by the electrical systems analysis team on a production line in a premium car factory in the interior of the state of Rio de Janeiro over a three-month period were taken into account, and the four most recurrent faults with the greatest impact on the factory's quality indicators were selected from a list of the "top ten". Once the unanalysed defects had been identified, an algorithm was developed to analyse each of them in detail with their respective outputs. Automating this analysis saves time because instead of searching for possible causes, as technicians do, this algorithm guides the user directly to the most likely cause. At the end of this application, the result is a saving of time in these analyses, which can last from hours to days and have an impact not only on quality indicators, but also on efficiency and the delivery of vehicles to customers.