> ML_DATASET // SAN-FRANCISCO-TRAFFIC-PEMS_v1.0
San Francisco Bay Area Highway Traffic (PeMS)
California Department of Transportation (Caltrans / PeMS) · Smart Transportation & Traffic Engineering · 17,544 hourly occupancy rate measurements across 863 highway loop sensors
Smart Transportation & Traffic EngineeringPublic Domain17,544 hourly occupancy rate measurements across 863 highway loop sensorsopen
Dataset Profile & Characteristics
Label Type:Road occupancy rate [0, 1] per sensor station
Languages:en
License Tier:permissive-open-source
Modalities:time-series, spatial
Intended Use
- Spatio-temporal graph neural network and deep forecasting benchmarking
Prohibited / Discouraged Use
- Single vehicle tracking or individual speed enforcement
Bias, Leakage & Privacy Risk Analysis
Privacy / Sensitive Data Risks:
Inductive loop electromagnetic road occupancy; no vehicle license or driver identities.
Known Bias:
Bay Area commute peaks (morning outbound, evening inbound) and weekend patterns.
Known Benchmark Leakage:
Highway graph adjacency provides spatial correlation across neighboring stations.