Traffic Congestion Pipeline¶
An open-source, installable Python pipeline for spatiotemporal traffic congestion analysis, integrating the HERE Traffic API, OSMnx, and PySAL.
Designed for reproducible urban analytics across Indonesian metropolitan cities: Semarang, Bandung, and Jakarta.
Features¶
- Eleven-stage analysis pipeline: aggregate → EDA → geostatistics → bottleneck → POI → synthesis → multilevel → markov → speed-validation → h3-robustness → positive-control
- CLI and Python API: Use
traffic-pipelinefrom the terminal or import modules directly - 316 million observations across 18,877 road segments over 12 months (Mar 2025–Mar 2026)
- OSM-based segment matching: Stable
osm_composite_idreplaces legacy FID approach - Reproducible: Zenodo-archived dataset + installable package + CI
Quick Start¶
See the Installation and Quick Start guides for details.
Cities Covered¶
| City | Segments | Coverage |
|---|---|---|
| Semarang | 1,076 | Urban core |
| Bandung | 3,063 | Metropolitan area |
| Jakarta | 14,609 | Greater Jakarta |
Data Availability¶
The aggregated dataset (24 GeoPackages — 8 time periods × 3 cities) is archived on Zenodo:
Hadi, F., Wahyuddin, Y., Sabri, L. M., & Indrajit, A. (2026). Traffic Congestion Dataset: Semarang, Bandung, Jakarta (Mar 2025–Mar 2026). Zenodo. doi:10.5281/zenodo.19211072
License¶
MIT — see LICENSE.