Bridge: Retrieval-Augmented Spatiotemporal Modeling for Urban Delivery Demand

TL;DR

Bridge introduces a retrieval-augmented spatiotemporal framework that enhances urban delivery demand forecasting, especially in cold-start regions lacking historical data, by combining a graph backbone with a time-aware memory and a future-aware retriever.

Contribution

The paper presents a novel retrieval-augmented model that improves demand forecasting in cold-start regions and transfer scenarios by integrating a memory module with a graph-based predictor.

Findings

Bridge outperforms baselines in cold-start and transfer scenarios.

Retrieval-augmented approach improves demand prediction accuracy.

Model effectively leverages operational memory for urban demand forecasting.

Abstract

Forecasting urban delivery demand becomes substantially more challenging when newly added service regions lack historical records. Existing spatiotemporal forecasters effectively model spatial dependence once sufficient node histories are available. Still, they remain parametric and therefore struggle to recover short-term operational dynamics in cold-start regions. Geospatial embeddings help identify where a region is and what function it serves, yet they do not directly reveal how a similar region behaves under a comparable temporal context. We propose Bridge, a retrieval-augmented spatiotemporal graph framework that combines an inductive contextual graph backbone with a time-aware memory of region-time windows. For each target region, Bridge retrieves future demand patterns from the memory using both regional context and recent dynamics, and refines the backbone forecast through a gated fusion mechanism. To align retrieval with forecasting utility, we further train the retriever with a future-aware objective that favors entries whose future trajectories best match the target. Experiments on four real-world delivery datasets show that Bridge consistently improves over competitive spatiotemporal baselines in both within-city cold-start and cross-city transfer with partial observations. The results show that retrieval augmentation provides a useful operational memory for cold-start urban demand forecasting when parametric graph generalization alone is insufficient.