隨著物聯網(Internet of Things， IoT)與測繪技術快速發展，「數位孿生(Digital Twins)」這個概念也逐漸在這個資訊化時代引起各領域關注。數位孿生能夠有效描述真實物理世界物件在虛擬資訊世界中的數位表徵，持續提供觀測對象的動態與靜態資訊，因而廣泛應用於工業4.0、智慧城市、自動化家居等物聯網相關應用。數位孿生作為物理世界與虛擬世界的連結，其資料若能以符合直覺的方式呈現於使用者所處環境，便有潛力帶來身歷其境且直觀的操作體驗 ；然而目前主流物聯網系統多仰賴於二維介面操作，必須仰賴使用者事先在腦中建立螢幕上的抽象標籤與三維物理環境中實體裝置的對應關係，是相當不直覺且低效率的 。此外，現行混合實境(Mixed Reality， MR)與物聯網(MR-IoT)系統及相關研究多採客製化介面，並將系統配置直接寫死(hard-coding)在底層程式碼中，導致新增裝置或場域遷移時皆需開發者重新介入，導致規模化部署及維護瓶頸。 針對上述問題，本研究提出一套以開放標準為核心的MR-IoT數位孿生整合架構，並設計標準化設定檔規格(Deployment Schema)作為部署資訊的結構化描述機制；整合Microsoft HoloLens 2混合實境裝置空間映射與全息影像渲染能力、Azure Spatial Anchors視覺定位機制，以及開放地理空間協會SensorThings API感測與致動標準，並基於上述基礎建構三個核心功能組件：空間映射與錨點定位、IoT標準化通訊、及使用者介面。透過此架構，單一混合實境應用得以透過國際開放標準介面存取跨廠牌異質物聯網裝置，並將控制介面以全息影像精確疊加於實體裝置旁，以消除使用者於二維螢幕與三維物理空間之間的認知轉換負擔。在設定檔規格層面，本研究提出基於JSON之設定檔格式，採用樣板與實例分離設計。該規格將裝置屬性、控制面板配置與空間繫結等資訊集中於結構化文件當中，並透過識別碼與Open Geospatial Consortium (OGC)所提出的SensorThings API標準對應，同時填補既有標準於混合實境應用部署的空缺，藉由修改設定檔取代複雜程式碼改寫維護，從設計層面大幅降低系統建置成本。 面對研究期間商用混合實境生態系的劇烈變動，本研究的設計亦印證了採用開放標準與跨平台架構的必要性。最終透過智慧家居情境建置原型系統，成功整合自製Arduino感測器與商用智慧燈泡，驗證所提架構的設備整合能力並討論設定檔部署之可行性。期望本研究能藉由降低部署成本與提升跨平台可移植性，為後續MR-IoT系統之異質裝置整合及跨場域規模化部署研究，奠定可靠之參考基礎。;With the rapid advancement of the Internet of Things (IoT) and geospatial technologies， the concept of Digital Twins has drawn growing attention across diverse fields in this information age. A digital twin effectively serves as the digital representation of real-world physical objects within a virtual information space， continuously providing both dynamic and static properties of the observed objects. Consequently， it has been widely adopted in IoT-related applications such as Industry 4.0， smart cities， and home automation. As a bridge between the physical and virtual worlds， digital twins hold the potential to deliver immersive and intuitive operational experiences—provided its data is presented intuitively within the user′s real-world environment. However， most mainstream IoT systems still rely on two-dimensional interfaces for interaction， requiring users to mentally establish mappings between abstract on-screen labels and physical devices in three-dimensional space—an approach that is both unintuitive and inefficient. Furthermore， existing Mixed Reality and IoT (MR-IoT) systems and related research commonly adopt custom interfaces with system configurations hard-coded directly into the underlying codebase. This means that adding new devices or migrating to a new venue requires repeated developer intervention， creating a significant bottleneck for scalable deployment and maintenance. To address these issues， this study proposes an open-standards-based MR-IoT Digital Twin integration architecture and introduces a standardized Deployment Schema as a structured mechanism for describing deployment information. The architecture integrates the spatial mapping and hologram rendering capabilities of Microsoft HoloLens 2， the visual positioning of Azure Spatial Anchors， and the sensing and actuation standards of the OGC SensorThings API (STA) to construct three core functional components: spatial mapping and anchor positioning， standardized IoT communication， and the user interface. Through this architecture， a single MR application can access heterogeneous IoT devices from different vendors via a unified international standard interface. Control panels are precisely overlaid on their corresponding physical devices as holograms， thereby eliminating the cognitive mapping burden imposed on users when bridging two-dimensional screens and three-dimensional physical space. At the deployment specification level， this study proposes a configuration file format based on JSON and JSON Schema， adopting a template-instance separation design. The Deployment Schema consolidates device attributes， control panel configurations， and spatial anchor bindings into a single structured document. It aligns with the STA standard through identifier references， filling the gap left by existing standards in MR application deployment. By enabling system updates through configuration file edits rather than complex code rewrites， the schema substantially reduces system construction and maintenance costs at the design level. In response to the significant shifts in the commercial MR ecosystem observed during the research period， the design approach taken in this study further affirms the necessity of adopting open standards and a cross-platform architecture. A prototype system was implemented in a smart home context， integrating a custom-built Arduino sensor and a commercial smart bulb， to demonstrate the device integration capability of the proposed architecture and to evaluate the feasibility of Deployment Schema-based deployment. It is hoped that this study， through reducing deployment costs and improving cross-platform portability， can provide a reliable reference foundation for future research on heterogeneous device integration and cross-venue scalable deployment in MR-IoT systems.
