Towards Trustworthy AI-Enabled Emergency Medical Dispatch: Uncertainty-Aware Ambulance Travel Time Estimation
Authors: Muhammad Adil Abid, Johan Holmgren, Fabian Lorig, Jesper Petersson
Organizations: Department of Computer Science and Media Technology, Malmö University, 21119 Malmö, Sweden · Sustainable Digitalisation Research Centre, Malmö University, Sweden · Department of Health Care Management, Region Skåne, 21428 Malmö, Sweden · Department of Neurology, Lund University, 22242 Malmö, Sweden
Emergency medical dispatch increasingly relies on AI-enabled decision support systems when making ambulance dispatch decisions. In this process, accurate ambulance travel time estimates are essential for timely resource allocation and prehospital care. However, existing ambulance travel time estimation approaches typically produce only single point predictions. Consequently, they do not indicate how reliable an individual estimate is or how much uncertainty it carries. This limitation is particularly problematic in safety-critical emergency medical service dispatch operations, where trips with similar expected travel times may exhibit substantially different levels of uncertainty. Furthermore, real-world ambulance trip logs may contain domain-specific artefacts, such as co-located healthcare facilities and physically implausible travel speeds. These artefacts can degrade point prediction accuracy and affect the prediction task on which subsequent uncertainty estimation depends. To address these challenges, we propose an uncertainty-aware methodology for ambulance travel time estimation and evaluate it through a retrospective real-world scenario study using two years of ambulance trip data from the Skåne region in Sweden. The proposed methodology moves beyond average-case point predictions by generating context-adaptive prediction intervals with stable empirical coverage across the evaluated operationally important contexts. The resulting prediction intervals provide explicit information about predictive uncertainty that can support more transparent and risk-sensitive ambulance dispatch decision-making.
Figures & tables
Figure 1: High-level workflow of the proposed uncertainty-aware ambulance travel time estimation methodology.
Figure 2: Step-by-step workflow of global split conformal prediction for uncertainty-aware ambulance travel time estimation.
Priority
Mean (min)
Std. Dev. (min)
CV
Trips
Priority 1 (Emergency)
29.79
13.84
0.465
2,345
Priority 2 (Urgent)
35.24
15.47
0.439
3,972
Priority 3 (Routine)
39.51
18.07
0.457
14,421
Table 1: Travel time variability across operational contexts.
Context
Coverage
All trips (average case)
96.5%
Morning peak (09:00)
94.4%
Rare routes (6–20 trips)
69.9%
Very rare routes (1–5 trips)
44.2%
Table 2: Empirical coverage of a naive global 95% reference interval computed from the training set mean and standard deviation.
Figure 3: Standard deviation of ambulance travel time by hour of day and day of week (Darker cells indicate greater travel time variability.)
Task formulation
MAPE ↓ (%)
MAE ↓ (min)
RMSE ↓ (min)
R2↑
Operationally naive (baseline)
12.47
4.83
10.86
0.62
Spatially disambiguated
12.90
4.65
10.65
0.65
Operationally valid
10.35
4.42
9.12
0.74
Table 3: Effect of domain-informed data curation on point prediction performance ( ↓ lower is better; ↑ higher is better).
Model
MAPE ↓ (%)
MAE ↓ (min)
RMSE ↓ (min)
R2↑
Polynomial Linear Regression
21.11
8.01
12.97
0.48
Random Forest
13.54
5.36
10.38
0.67
CatBoost one-hot
10.37
4.45
9.57
0.72
CatBoost native
10.36
4.44
9.42
0.73
ANN
10.35
4.42
9.12
0.74
Table 4: Comparison of candidate point prediction models ( ↓ lower is better; ↑ higher is better).
Method
Coverage (%)
Mean width ↓ (min)
Median width ↓ (min)
Baselines
Gaussian ( f(x)±1.645σ^ )
93.2
24.7
24.7
Quantile regression (uncalibrated)
89.7
20.4
19.3
Conformal
Global split conformal
89.1
17.6
17.6
CQR
90.5
20.8
19.8
Mondrian (priority-stratified)
89.0
17.6
18.0
Table 5: Comparison of uncertainty estimation methods at nominal 90% coverage ( ↓ lower is better).
Figure 4: Calibration of empirical versus nominal coverage. Dashed line: perfect calibration; markers: nominal 90% coverage.
Figure 5: Empirical coverage and interval width across operational contexts at nominal 90% coverage.
Accurate ambulance travel time estimation is critical for emergency medical dispatch, yet existing decision-support practices often rely on static lookup tables or simplified assumptions regarding how ambulance travel time differs from ordinary passenger-car travel. For instance, they assume ambulances simply travel a fixed percentage (e.g., 5%) faster than ordinary passenger cars. However, such assumptions are rarely validated using real ambulance operational data and may fail to capture the effects of trip priority, temporal context, and origin-destination. We address these limitations using historical ambulance trips from the Skåne region in Sweden. Our analysis reveals that ambulance travel times vary substantially by hour, day, and trip priority, and that commercial navigation systems fail to capture emergency-specific dynamics. We contribute a machine learning system to more accurately estimate ambulance travel time. To make the system interpretable, we explored SHapley Additive exPlanations and permutation importance techniques to ensure that our system captures operationally meaningful patterns.
Muhammad Adil Abid, Johan Holmgren, Fabian Lorig +1
Department of Computer Science and Media Technology, Malmö University, 21119 Malmö, Sweden · Sustainable Digitalisation Research Centre, Malmö University, Sweden · Department of Health Care Management, Region Skåne, 21428 Malmö, Sweden +1
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.
Zikun Lin, Daniel Zhuoyu Long, Viet Anh Nguyen
Department of Systems Engineering and Engineering Management, The Chinese University of Hong Kong
Artificial intelligence has strong potential to support clinical decision-making, yet its adoption in healthcare remains limited due to a lack of trust. Uncertainty estimation can signal unreliable predictions, and explainable AI (XAI) can clarify how predictions are made but existing methods treat them separately, providing no feature-level insight into why a prediction is uncertain or which tests to prioritize to reduce it. To address this gap, we propose explainable uncertainty estimation, which unifies uncertainty estimation and XAI to both quantify uncertainty and explain feature-level contributions. We introduce the Expected Gradients Reconstruction Uncertainty Estimate (egRUE), which incorporates prediction explanations into its uncertainty computation and decomposes uncertainty into feature-wise contributions. We prove theoretical properties of egRUE and show through experiments that it improves reliability and interpretability compared to existing methods. A user study with medical experts further demonstrates that egRUE's explanations improve calibrated trust over uncertainty scores alone, increasing confidence in correct predictions and reducing confidence in incorrect ones. By combining prediction uncertainty with feature-level explanations, egRUE strengthens decision-making support in safety-critical healthcare settings, clarifying both when predictions may be unreliable and which features drive that uncertainty.
Li Rong Wang, Jamie Duell, Xinran Xu +6
College of Computing and Data Science (CCDS) Nanyang Technological University (NTU) A*STAR Centre for Frontier AI Research Singapore · School Of Computing and Digital Technologies Sheffield Hallam University South Yorkshire, England · CCDS NTU Singapore +5