Dataset Card for Dataset Name

This dataset simulates realistic ride hailing demand patterns in a university campus environment by transforming real-world bike sharing trip data into ride hailing scenarios. The synthetic generation preserves temporal demand patterns and spatial distributions while adapting the characteristics for ride hailing services. This makes it valuable for transportation research, demand forecasting, and mobility optimization algorithms without privacy concerns.

Dataset Details

Direct Use

Useful for Reinforcement Learning applicaitons in mobility on demand, autonomous mobility on Demand, ride sharing and ride hailing applications. Used in Deep Reinforcement Leanring pre-trainig for decision making of autonomous agents.

Dataset Structure

TBA

Dataset Creation

Data Collection and Processing

See the corresponding paper [Erduran et al. 2019]

Who are the source data producers?

Data set sourced from German bike-sharing provider "Deutsche Bahn Call a Bike" (callabike.de). Data processing and transformation of geo-spatial information based on [Erduran et al. 2019].

It has been open-sourced via Cc 4.0 license.

[More Information Needed]

Citation

BibTeX:

@inproceedings{erduran2019multi, title={Multi-agent learning for energy-aware placement of autonomous vehicles}, author={Erduran, {"O}mer Ibrahim and Minor, Mirjam and Hedrich, Lars and Tarraf, Ahmad and Ruehl, Frederik and Schroth, Hans}, booktitle={2019 18th IEEE International Conference On Machine Learning And Applications (ICMLA)}, pages={1671--1678}, year={2019}, organization={IEEE} }

Dataset Card Authors and Contact

oemer450