Comprehensive spatiotemporal dataset of shared bicycle operations in Xiamen China

We present a comprehensive spatiotemporal dataset of shared bicycle operations in Xiamen, China, collected over five consecutive weekdays (December 21-25, 2020) during morning peak hours (6:00-10:00 AM). The dataset comprises three components: (1) 220,675 origin-destination trip records from 53,630 unique bicycles, including precise GPS coordinates, timestamps, and lock status; (2) 2,849,243 GPS trajectory tracking points captured at 30-second to 2-minute intervals during complete trips; (3) 14,071 electronic fence records with unique identifiers and four-corner coordinates specifying parking zone boundaries. The dataset captures environmental variations including weather impacts and maintains strict privacy protection through cryptographic anonymization. This open dataset enables analysis of urban mobility patterns, cycling behavior, and demand forecasting. We demonstrate the dataset's utility through temporal distribution analysis, spatial pattern visualization, regional clustering, and demand prediction using neural networks. The dataset supports research in urban transportation planning, shared mobility optimization, and smart city applications.