PROJECT

AURA

OpenTrainingData for Automated and Autonomous Micromobility

Problem Statement

Micromobility can reduce CO2 emissions and traffic. Previous developments in automated driving have been heavily focused on motor vehicles. Therefore, there is currently no open dataset that enables the development of autonomous driving functions in the context of micromobility away from motor vehicle lanes. As a result, early-stage developer teams and universities find it very difficult to develop initial procedures and train AI/ML applications.

Project Objective

The goal of OPTmicro is to collect, process, and provide an open training dataset for automated and autonomous micromobility to the scientific community and businesses. This will provide mobility solution developers with a dataset freely available as OpenData and Open Source, tailored to application-specific requirements such as driving away from motor vehicle lanes.

Implementation

The project includes equipping and modifying an existing test vehicle as well as designing a data collection strategy. This will involve collecting data in various traffic situations in different operational areas and time periods. The data will be processed manually and semi-automatically and published as an open dataset. In addition, based on the dataset, AI models will be trained as references, which will also be made publicly accessible.

Project Partners

• Merseburg University of Applied Sciences
• FusionSystems GmbH