Whereas material data are used everywhere at the core of any engineering design, physical material testing is very expensive and time consuming. Moreover, the number of experimental tests rapidly increases with the different combinations of material systems, environmental conditions, load case scenarios etc. These engineering material data can be leveraged to further enrich material databases. The proposed workflow is based on a reduced set of existing material data that is combined with machine learning and cloud computing technologies to accelerate the generation of material data at minimal cost. The data enrichment can be performed using three methodologies. Firstly, a pure data science approach can be adopted. Secondly, a physics-based approach can be adopted. Thirdly, a mixed approach can be adopted by combining both the powers of the pure data science and the physics-based approaches. To illustrate the above-described workflow, a case study is presented. It reports an end-to-end virtual manufacturing solution, that starts from the process modeling stage, and ends at assessing a quality index of the final manufactured part, by going through considering different sources of uncertainties, focusing on the used material and the surrounding environmental conditions. In this framework, the proposed workflow is used to feed the uncertainty-based design study by the needed material cards, considering the different sources of uncertainties. More specifically, the presented case study focuses on reinforced plastics for the automotive industry. It exhibits a full demonstration about how to leverage smart material data generation using materials informatics, for optimizing probabilistically, key structural performances of parts in electric vehicles, made of these materials. The proposed work is concluded by assessing objectively the gains achieved by such an end-to-end virtual solution for an automotive engineering team. The gain is assessed from 3 perspectives: Time, Budget, and Carbon Footprint.