The AutoMat CVIM - A Scalable Data Model for Automotive Big Data Marketplaces

TL;DR

This paper presents AutoMat CVIM, a scalable, brand-independent data model for automotive big data marketplaces, enabling efficient crowd-sourced sensor data utilization and traffic analysis.

Contribution

It introduces the Common Vehicle Information Model (CVIM), a harmonized data model for vehicle sensor data, facilitating cross-brand data integration in automotive big data platforms.

Findings

Demonstrates a vehicle big data processing pipeline using CVIM.

Shows aggregation and evaluation of fleet sensor data for traffic measurement.

Validates the scalability and effectiveness of the data model.

Abstract

In the past years, connectivity has been introduced in automotive production series, enabling vehicles as highly mobile Internet of Things sensors and participants. The Horizon 2020 research project AutoMat addressed this scenario by building a vehicle big data marketplace in order to leverage and exploit crowd-sourced sensor data, a so far unexcavated treasure. As part of this project the Common Vehicle Information Model (CVIM) as harmonized data model has been developed. The CVIM allows a common understanding and generic representation, brand-independent throughout the whole data value and processing chain. The demonstrator consists of CVIM vehicle sensor data, which runs through the different components of the whole AutoMat vehicle big data processing pipeline. Finally, at the example of a traffic measurement service the data of a whole vehicle fleet is aggregated and evaluated.