環保專案成果報告查詢系統

固體再生燃料管理系統發展先期規劃

專案計畫編號
經費年度	113
計畫經費	1,460 千元
預算科目	資源循環管理-資源循環管理
專案開始日期	2024/10/16
專案結束日期	2024/12/30
專案主持人	蔡明璋
主辦單位	循環署循環處理組
承辦人	楊湘雲
執行單位	逢甲大學
專案分類	廢棄物管理
中文關鍵字	固體再生燃料，儀表板規劃設計，人工智慧
英文關鍵字	Solid Recovered Fuel, Dashboard planning and design, Artificial Intelligence
協同主持人
共同主持人
計畫聯絡人	傅以恩
計畫聯絡電話	04-24516609#330
計畫聯絡信箱	winnifu@gis.tw

中文摘要	本計畫以規劃「固體再生燃料（SRF）管理系統」為核心，旨在透過系統化功能模組設計，提供循環署、地方環保局等單位掌握SRF之製造與使用運作情形。執行內容包括系統功能規劃與成本評估、儀表板雛型設計、雲端架構評估以及AI影像辨識技術與空污排放推估技術之發展規劃。成果顯示，本計畫規劃之系統功能模組設計內容涵蓋首頁、製造與使用廠管理平台、運作儀表板等，能透過直觀數據圖表展示各時期或地區之SRF的製造量、使用量，同時也可了解SRF製造廠料源廢棄物類別比例，以及SRF使用廠類別比例，後續計畫進行開發建置，上線後可提供互動式查詢管理平台，作為SRF管理最重要之平台。在導入AI技術方面，建議從外形較易辨識的原料（如PVC管）著手發展料源辨識技術，逐步累積影像庫並提升模型準確性；空污排放推估則需建立SRF來源與排放數據的關聯，未來可針對燃料比例或鍋爐操作提供改善建議。短期內建議推動CCTV影像標準串流規格，並配合法規修訂調整系統，中長期則需強化數據庫建設，提升產銷資料關聯性與分析能力，為SRF管理系統的未來發展奠定基礎。
中文關鍵字	固體再生燃料，儀表板規劃設計，人工智慧

中文摘要

本計畫以規劃「固體再生燃料（SRF）管理系統」為核心，旨在透過系統化功能模組設計，提供循環署、地方環保局等單位掌握SRF之製造與使用運作情形。執行內容包括系統功能規劃與成本評估、儀表板雛型設計、雲端架構評估以及AI影像辨識技術與空污排放推估技術之發展規劃。成果顯示，本計畫規劃之系統功能模組設計內容涵蓋首頁、製造與使用廠管理平台、運作儀表板等，能透過直觀數據圖表展示各時期或地區之SRF的製造量、使用量，同時也可了解SRF製造廠料源廢棄物類別比例，以及SRF使用廠類別比例，後續計畫進行開發建置，上線後可提供互動式查詢管理平台，作為SRF管理最重要之平台。在導入AI技術方面，建議從外形較易辨識的原料（如PVC管）著手發展料源辨識技術，逐步累積影像庫並提升模型準確性；空污排放推估則需建立SRF來源與排放數據的關聯，未來可針對燃料比例或鍋爐操作提供改善建議。短期內建議推動CCTV影像標準串流規格，並配合法規修訂調整系統，中長期則需強化數據庫建設，提升產銷資料關聯性與分析能力，為SRF管理系統的未來發展奠定基礎。

中文關鍵字

固體再生燃料，儀表板規劃設計，人工智慧

類型	檔名	檔案大小	說明	最新版公開日期
期末報告	113SRF_成果報告0103_final.pdf	11MB		2025/3/7 上午 12:00:00

摘要	This project focuses on planning the “Solid Recovered Fuel (SRF) Management System,” aiming to design systematic functional modules to assist organizations such as the Environmental Protection Administration (EPA) and local environmental bureaus in monitoring the production and usage of SRF. The scope of work includes system functionality planning and cost evaluation, dashboard prototype design, cloud architecture assessment, and the development of AI-based image recognition and air pollution emission estimation technologies. The project outcomes reveal that the planned system includes functional modules such as a homepage, manufacturing and usage plant management platforms, and an operational dashboard. These modules enable users to view production and usage volumes of SRF across different periods or regions through intuitive graphical data representations. They also provide insights into the waste material categories used by SRF manufacturers and the category proportions of SRF users. Upon completion of system development and launch in subsequent phases, this interactive query and management platform will serve as the key tool for SRF management. For AI technology integration, it is recommended to start with material source recognition technology focusing on easily identifiable materials, such as PVC pipes. By gradually building an image database and improving model accuracy, the system can enhance material identification capabilities. Additionally, establishing correlations between SRF sources and emissions data will allow for air pollution emission estimation, offering recommendations for fuel ratio adjustments or boiler operation optimization in the future. In the short term, it is recommended to promote standardized streaming specifications for CCTV integration and align system adjustments with regulatory revisions. For the long term, strengthening database construction and enhancing the correlation and analytical capabilities of production and sales data will lay a solid foundation for the future development of the SRF management system.
英文關鍵字	Solid Recovered Fuel, Dashboard planning and design, Artificial Intelligence

摘要

This project focuses on planning the “Solid Recovered Fuel (SRF) Management System,” aiming to design systematic functional modules to assist organizations such as the Environmental Protection Administration (EPA) and local environmental bureaus in monitoring the production and usage of SRF. The scope of work includes system functionality planning and cost evaluation, dashboard prototype design, cloud architecture assessment, and the development of AI-based image recognition and air pollution emission estimation technologies. The project outcomes reveal that the planned system includes functional modules such as a homepage, manufacturing and usage plant management platforms, and an operational dashboard. These modules enable users to view production and usage volumes of SRF across different periods or regions through intuitive graphical data representations. They also provide insights into the waste material categories used by SRF manufacturers and the category proportions of SRF users. Upon completion of system development and launch in subsequent phases, this interactive query and management platform will serve as the key tool for SRF management. For AI technology integration, it is recommended to start with material source recognition technology focusing on easily identifiable materials, such as PVC pipes. By gradually building an image database and improving model accuracy, the system can enhance material identification capabilities. Additionally, establishing correlations between SRF sources and emissions data will allow for air pollution emission estimation, offering recommendations for fuel ratio adjustments or boiler operation optimization in the future. In the short term, it is recommended to promote standardized streaming specifications for CCTV integration and align system adjustments with regulatory revisions. For the long term, strengthening database construction and enhancing the correlation and analytical capabilities of production and sales data will lay a solid foundation for the future development of the SRF management system.

英文關鍵字

Solid Recovered Fuel, Dashboard planning and design, Artificial Intelligence

相關專案最新公開

112年度臺南市推動固體再生燃料...

112年度嘉義縣環境保護局AI人...

112年環境督察總隊計畫管理支援...

113年度臺南市推動固體再生燃料...

水污染源管制資料管理系統優化計畫...