Towards Reliable Zero-Shot Crowd Forecasting: Evaluating Time Series Foundation Models for Special Event Pedestrian Forecasting
2026-07-20 • Machine Learning
Machine Learning
AI summaryⓘ
The authors study how to predict crowd movements during rare special events, which is hard because there isn't much past data and events are short. They explore using pretrained time series models to forecast crowd flow without needing to retrain models specifically for each event. Their approach also estimates how uncertain these predictions are, which helps with planning for sudden changes or risks. They test their methods on a real event and provide guidance on when these quick predictions can be trusted by crowd managers.
pedestrian-flow forecastingprobabilistic uncertainty quantificationtime series foundation modelszero-shot forecastingshort event observation windowscrowd managementpredictive uncertaintytail risks
Authors
Ziteng Li, Yanan Xin, Tina Comes, Serge Hoogendoorn
Abstract
Managing massive crowds during infrequent special events requires reliable real-time pedestrian-flow forecasting to ensure public safety and operational efficiency. However, supervised forecasting methods face limitations in these contexts due to scarce historical data, heterogeneous data distributions, and short in-event observation windows. To effectively support operational decision-making, forecasts should provide not only accurate point estimates but also informative predictive uncertainty. Probabilistic uncertainty quantification plays a critical role in this aspect, particularly capturing sudden volatility and tail risks. This paper investigates pretrained time series foundation models as a lightweight approach for zero-shot probabilistic forecasting without extensive local retraining. Using decision-oriented metrics tailored to short events, we conduct a comprehensive assessment of two time series foundation models on crowd forecasting, with the SAIL2025 event as a use case. We then distill practical insights for crowd managers, specifying when zero-shot forecasts remain operationally reliable.