Contextual Data Integration for Bike-sharing Demand Prediction with Graph Neural Networks in Degraded Weather Conditions

TL;DR

This work tackles predicting bike-sharing Origin–Destination demand under degraded weather by extending a graph-based predictive framework (ST-ED-RMGC) with contextual data. It introduces an embedding module for calendar structure, adds weather and road-car flow information, and employs a time-embedding mechanism to capture temporal dependencies, improving robustness to adverse conditions. In Lyon, weather features significantly reduce forecasting errors in rainy scenarios (over 20% improvement in MSE) and time embedding yields consistent gains across conditions, with additional benefits when incorporating car-flow data in certain cases. The results support multi-modal contextual cues as a practical strategy to enhance OD-based bike-sharing forecasting and to inform operations during degraded weather.

Abstract

Demand for bike sharing is impacted by various factors, such as weather conditions, events, and the availability of other transportation modes. This impact remains elusive due to the complex interdependence of these factors or locationrelated user behavior variations. It is also not clear which factor is additional information which are not already contained in the historical demand. Intermodal dependencies between bike-sharing and other modes are also underexplored, and the value of this information has not been studied in degraded situations. The proposed study analyzes the impact of adding contextual data, such as weather, time embedding, and road traffic flow, to predict bike-sharing Origin-Destination (OD) flows in atypical weather situations Our study highlights a mild relationship between prediction quality of bike-sharing demand and road traffic flow, while the introduced time embedding allows outperforming state-of-the-art results, particularly in the case of degraded weather conditions. Including weather data as an additional input further improves our model with respect to the basic ST-ED-RMGC prediction model by reducing of more than 20% the prediction error in degraded weather condition.

Figures (4)

• Figure 1: Framework of the ST-ED-RMGC model, adapted from KE2021102858
• Figure 2: Embedding module. The Dense Module is composed of 3 dense layers in series.
• Figure 3: Integration of the embedding module into the existing architecture.
• Figure 4: IRIS zones without any aggregation (dashed line) and the aggregation into 50 zones (solid line).