公共管線圖資更新作業之常見錯誤與改善建議

Common Mistakes and Improvement Suggestions in Updating the Public Pipeline Maps

在管線資料進行建置與維護的過程中，新埋或換埋之管線須由管線單位辦理圖資更新作業，其管線空間位置之正確性以及屬性資料之完整性能否加以確保，將與管線資料庫的品質永續及其運用成效息息相關。為使管線圖資更新資料能夠詳實加以填報，內政部國土管理署於2024年首次規畫執行了與該項作業相關之品質稽核作業。由資料完整性抽查及正確性抽測二個層面之執行成果發現，完整性錯誤發生之分類占比以未提交完整資料（A類）的68%為最多，最常出現的錯誤即為埋深佐證照片的缺漏；另在正確性錯誤之分類占比上，則以高程坐標精度不合格（D類）的59%較高。針對此次稽核作業所發掘之通用性問題，可分別在埋深作業類提出3項、定位測量作業類提出2項、地理標記語言GML（Geography Markup Language）填報作業類提出6項以及管理作業類提出3項，合計14項的問題解決相關建議，當可做為中央與地方政府在推動管線資料庫建置及提升管線資料品質相關工作之參考。

During the process of building and maintaining pipeline data, newly buried or replaced pipelines must be updated by the pipeline-own company. Whether the accuracy of the pipeline's spatial location and the integrity of its attribute data can be ensured, is closely related to the sustainability and application of the pipeline database. In order to ensure the pipeline map updated data can be submitted in detail, the National Land Management Administration of the Interior Ministry planned and implemented audit work related to the quality of this operation for the first time in 2024. From the results of the implementation of data integrity (inspection) and data accuracy (positioning), it was found that the highest proportion of integrity errors was 68% for failure to submit complete data (Class A), and the most common error was the omission of depth supporting photos; in terms of the proportion of accuracy errors, the highest proportion was 59% for unqualified height measurements (Class D). In response to the common problems discovered during this audit, 3 suggestions in the depth operation category, 2 suggestions in the positioning operation category, 6 suggestions in the GML (Geography Markup Language) reporting operation category, and 3 suggestions in the management operation category, totaling 14 problem-solving suggestions, were proposed. These suggestions can be used as a reference for the central and local governments in promoting the establishment of pipeline databases and improving the quality of pipeline data.