Search papers, labs, and topics across Lattice.
This paper introduces HERALD, a unified deep learning framework that forecasts the emergence and evolution of traffic crash hotspots by integrating historical crash data into weekly risk maps. Utilizing a CNN-Transformer architecture with a mixture-of-experts approach, HERALD effectively adapts to both dense urban and sparse rural environments, providing timely warnings of new hotspots before they manifest. The framework outperforms five baseline models in accuracy and precision across six counties in Wisconsin, fundamentally transforming hotspot management from reactive to proactive strategies.
HERALD shifts traffic safety from reactive mapping of past crashes to proactive forecasting of emerging hotspots, enabling timely interventions.
Road crashes remain among the gravest threats to public safety, and preventing them is a defining task of transportation systems worldwide. Much of that harm concentrates at hotspots, yet a hotspot is less a place than an episode; it emerges quietly at an intersection or along an arterial, intensifies for weeks, then subsides, only to reappear elsewhere. Enforcement guided by maps of past crashes inevitably trails this cycle, patrolling yesterday's hotspots while tomorrow's form unwatched. Breaking that lag requires three capabilities at once: detecting hotspots as they are born, forecasting where they will sit next week, and following each one through its life. We introduce HERALD (Hotspot Emergence, Risk Anticipation, and Life-cycle Dynamics), a unified deep learning framework that provides all three from a single statewide model. HERALD distills each county's recent crash history into weekly risk maps and forecasts the next with a CNN--Transformer, whose mixture-of-experts lets one model serve dense urban cores and sparse rural corridors alike. Each forecast is anchored in the county's long-run crash geography, sharpened by the self-exciting effect of recent crashes, and paired with explicit warnings of where new hotspots are about to appear. Followed over time, every hotspot acquires a legible life story, from birth through growth and stability to decline and death. Across six heterogeneous Wisconsin counties, HERALD forecasts more accurately than five identically trained baselines, locates hotspots most precisely, and flags emerging risks before they take hold. A single adjustable setting trades accuracy for extra sensitivity where deployment demands it. The result shifts hotspot management from mapping the past to anticipating the future.
