The recent technological advancement in the connected space of the automobile sector like improved connectivity and integration of 5g technologies besides government norms for compulsory telematics is pushing the automotive industry towards a new age of data. The abundance of vehicle operational data like parameter data from various sensors such as coolant temperature, ambient pressure, demand torque, engine rpm, etc., fault codes, and external devices data like TPMS. Which flows on a near real-time basis through telematics and has paved the path for IoT concepts like digital twin in Automotive Industry. A digital twin in the automotive industry may be a cloud-based real-time digital counterpart of a physical entity like an entire car, mechanics, components, electrics, software, or the physical behavior of a vehicle. The data that flows from the physical product to the digital/virtual model is the connections between the digital world and the physical environment. The digital twin can also represent the operating profile based on historical data like service history, Job card details, configuration changes, parts replacement, and warranty data depending on its intended purpose. The digital twin model can be categorized based on the availability of this data in, i) a data-driven model. ii) physics-based model. These digital twin models can serve different purposes like product testing, performance optimization through design changes, improve monitoring capabilities of vehicle fleets, and predictive maintenance. This paper illustrates the design of a data-driven model of a turbocharger a critical automotive component and explains the data mining and processing strategies in the digital twin model for failure prognostic of a component to predict the failure through parametric and nonparametric machine learning algorithms. And establishing counteractions for predicted failures.