Abstract
In this paper, we present an implementation of a research data management system that features structured data storage for spatio-temporal experimental data (environmental perception and navigation in the framework of autonomous driving), including metadata management and interfaces for visualization and parallel processing. The demands of the research environment, the design of the system, the organization of the data storage, and computational hardware as well as structures and processes related to data collection, preparation, annotation, and storage are described in detail. We provide examples for the handling of datasets, explaining the required data preparation steps for data storage as well as benefits when using the data in the context of scientific tasks.
Highlights
There is a growing awareness in the research community of the importance of FAIR principles in data handling [1]: data should be free, accessible, interoperable, and reusable
In this paper, we present an implementation of a research data management system that features structured data storage for spatio-temporal experimental data, including metadata management and interfaces for visualization and parallel processing
In order to allow for a scalable storage and multi-user access, the data are automatically imported into a spatial database ( [43]), where all metadata of the datasets are directly accessible for complex queries
Summary
There is a growing awareness in the research community of the importance of FAIR principles in data handling [1]: data should be free, accessible, interoperable, and reusable. In order to generate a realistic representation of the dynamic situation at the time of data capture, the data have to be integrated in a holistic data management system This system allows for conducting seamless experiments with arbitrary sensor combinations based on the stored data. The established uniform storage and documentation schema should be transformable into the target formats of data publication platforms to support re-use by other researchers In this contribution, we report on our realization of a data management system for large, heterogeneous spatio-temporal datasets that suit our requirements: Sensor data are stored in a structured, well-documented, and interoperable way. Derived datasets produced by the analyses of these individual research projects are stored in addition to the raw sensor data This allows more complex analyses in the domain of (real-time) positioning, making use of higher-level knowledge. This aspect, is beyond the scope of this paper
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
