Data analytics methodology for vehicle quality complaint analysis

Automobiles have become a part of people's everyday life. Customers base their trust on a brand's reputation while making a buying decision. With the evolving trends, feature advancements, automobiles have become a complex system. Ensuring product quality amidst system complexities is necessary to maintain the brand value & ensure customer satisfaction. Efficient addressal of customer quality complaints is pivotal in maintaining product quality. Until recent times, manual analysis using traditional data processing tools has been used for vehicle complaint analysis. However, considering the large quantum of data in new age vehicles with more number of control units, it becomes time-consuming, expertise dependent and susceptible to human errors. Given this & the advent of new age machine learning techniques & data analysis tools that can handle large amount of data, it is imperative to adopt them for vehicle complaint analysis. This research paper puts forth a method to analyze vehicle complaints by identifying & extracting important features from a raw dataset, representing them in a computable form, calculating statistical relationship metric and establishing a meaningful cluster that can help the engineering team to address the complaint effectively. The results show that based on the proposed method, it is possible to identify attributes & their combinations that could be potential cause or effect of a defect. This helps to fasten the root cause identification and solution definition in a data driven manner.