The spread of new energy vehicles has opened a new chapter in premium-class acoustics. In e-mode, hitherto inaudible noises - suppressed by traditional engines - have become audible. Therefore, it is necessary to make subjective noises objective. The first step is to make them comparable by measuring. The result of the sum of these spectrograms is an acoustic fingerprint of the vehicle. The next step is to add them to a database. Early failure prediction and the avoidance of redundant service activities by using automated event detection led to financial savings. Data processing with machine learning opens space for car workshop support applications, and it is also possible to prevent critical failures that may begin with the help of predictive models (predictive service).