With increasing awareness on environmental issues, there is a necessity to make the environmental performance of a product assessable. The most valid and scientifically recognized methodology to achieve this target is the Life Cycle Assessment (LCA). In the automotive industry, the carbon footprint is the main focus when it comes to the environmental performance. A vehicle consists of many components. This makes an overall LCA of a vehicle a complex project, which cannot be done without the support of databases and software tools. One possibility to handle the complexity is the assignment of components to pre-defined Life Cycle Inventory (LCI) models based on secondary data, which can lead to inaccuracies if the assignment is not done by the most influencing component attributes. This paper tackles the conflict of inaccuracy and modelling effort of the automated LCAs with assigned LCI models. A methodology to build a construct of generic LCI models under constraints regarding accuracy and modelling effort is here presented and discussed throughout an example. For this approach, the relevant factors influencing the Global Warming Potential (GWP) of the component have to be identified. A distinction has to be made as to whether the influencing factors are described by discrete or continuously variable parameters. Considering the constraints on accuracy and modelling effort, the parameters have to be discretized and grouped to meaningful grid points. The number of grid points can be reduced when technical constraints are considered. The method is applied on the cradle-to-gate modelling of aluminium automotive components for the main technologies casting, deep-drawn sheets and extrusion. Due to a systematic construct of pre-defined LCI models, the assignment of LCI models is more flexible and the effect of the most relevant component attributes on the GWP are observable.