Selective Ambulance Dispatch Under Contextual Travel-Time Uncertainty

2026-05-22

AI Summary

This study addresses the challenge of ambulance dispatch for out-of-hospital cardiac arrest (OHCA) emergencies, where timely response must be balanced against limited resources under dynamic traffic conditions that static zoning and deterministic travel-time models fail to capture. The authors propose the IDEAL framework, which leverages contextual learning to estimate segment-level travel times and triggers dual-dispatch only when the optimistic time gap between primary and backup routes exceeds a learned threshold. Innovatively, it models travel-time uncertainty via Burg divergence-based perturbations, inferring context-aware radii from historical underestimation errors. The resulting optimistic gap computation is cast as a difference-of-convex program, enabling efficient real-time decision-making. Evaluated on real OHCA data from the Hong Kong Fire Services Department with real-time simulation, IDEAL significantly outperforms baseline approaches—including fixed zoning and Google Maps routing—achieving a superior trade-off between response time and resource utilization.

Abstract

Ambulance response is time-critical in out-of-hospital cardiac arrest (OHCA), where dispatchers must balance timely arrivals with limited fleet capacity. Static territories and deterministic travel-time estimates are vulnerable to dynamic congestion, while always-dual dispatch adds redundancy but consumes fleet capacity. We propose IDEAL (Intelligent Dual dispatch of Emergency AmbuLances), a selective dual-dispatch framework that sends a second ambulance only when the optimistic gap between primary and secondary paths exceeds a threshold. IDEAL learns context-specific edge travel times from trip-level dispatch records, including unobserved routes, using a weakly supervised bilevel representation network. We train the nonsmooth model with mini-batch conservative gradients and prove an asymptotic convergence guarantee. IDEAL models uncertainty via Burg-divergence perturbations to a shared metric in the learned representation space, thereby inducing correlated changes in edge travel times and learning context-specific radii from historical underprediction errors. For real-time decisions, IDEAL casts optimistic-gap computation as a difference-of-convex program and derives an efficient oracle with complexity guarantees. In collaboration with the Hong Kong Fire Services Department, we evaluate IDEAL using historical OHCA records and real-time adaptive simulations. The results achieve a stronger response-time/resource trade-off relative to all region-based and Google-based baselines.

Problem

• Ambulance Dispatch
• Travel-Time Uncertainty
• Out-of-Hospital Cardiac Arrest
• Resource Allocation
• Emergency Response

Innovation

• selective dual dispatch
• contextual travel-time uncertainty
• weakly supervised bilevel representation
• Burg-divergence perturbation
• difference-of-convex optimization