Long-Horizon Traffic Forecasting via Incident-Aware Conformal Spatio-Temporal Transformers

Mayur Patil, Qadeer Ahmed, Shawn Midlam-Mohler, Stephanie Marik, Allen Sheldon et al.|March 17, 2026arXiv

Key Takeaway

Incorporating incident severity signals and dynamic road relationships into spatio-temporal models significantly improves long-horizon traffic predictions with calibrated confidence intervals—practical for real-world transportation planning.

Summary

This paper improves traffic forecasting by using a Transformer model that understands both spatial patterns (how traffic flows across roads) and temporal patterns (how it changes over time), while accounting for incidents like crashes.

reasoning evaluation applications

Key Terms

spatio-temporal-reasoning conformal-prediction dynamic-graph-construction monte-carlo-simulation coefficient-of-variation