本研究利用公共自行車系統分站車位數與分時可用車輛數，引進風險管理 概念，建立缺車、缺位、可靠度、波動率、平均等候時間、使用率等六大營運 服務指標。依據連續半年臺北市YouBike 租賃站每五分鐘時階車輛數資訊，針 對逾808 萬筆有效數據進行分析得出：(1) 辨識五類空間熱點及三類時間熱點； (2) 探討平日與假日營運特性異同，進而建議後續應評估實施差別費率與備援 車輛機制；(3) 釐清影響租賃站狀態的地理因素，以輔助現行設站準則；(4) 實 例試算鄰站支援缺車熱點之效果，提出使用者主動式調度的具體作法。本研究 將有限的原始資料發揮多元應用，指標模型可適用各公共自行車系統供主管與 營運單位參考，建議未來能有更多元的開放式數據，作為共享運輸創新的基石。

This study adopts chronic data of public bike systems regarding the numbers of available bikes and spaces of each station. Six operational indicators were developed in light of risk management, including the risk of insufficient bikes, risk of insufficient spaces, reliability, fluctuation, waiting time, and utilization rate. We collected Taipei’s “YouBike” data with a time span of 6 months for every 5 min., or a data size of 8,080,128. The findings include: (1) identification of five types of spatial hotspots and three types of temporal hot periods; (2) clarification of the operational characteristics on weekdays and weekends, upon which differential pricing and backup mechanism were based; (3) proposal of geographic factors that would affect the status of bike stations－a complement to the current criteria for building new stations; and (4) quantification of the joint effect of adjacent stations to ease hotspots along with an approach to active bike dispatch. This study demonstrated multiple applications given very limited data types. The authorities can use the model for better operations management. It is suggested that more open data to be released for innovative shared transport.