Accurate digital representation of sensor locations is essential for the creation of operational digital twins in smart buildings. This paper introduces a geometry-driven method for automatic digital placement of IoT sensors within Building Information Models (BIMs), requiring no semantic metadata or manual annotation. The method analyses the three-dimensional geometry of rooms, walls, and openings to infer plausible and rule-based sensor positions that reflect real-world installation practices. Unlike optimization-based approaches that seek coverage or cost efficiency, the present work focuses on geometry-only reasoning for visual and spatial alignment between physical sensors and their digital counterparts. The algorithm is implemented in a browser-executable environment using three.js, enabling real-time visualization of sensor layouts directly from mesh-based BIMs. Validation in a live office building with more than 300 sensors demonstrates that the computed placements reproduce actual installation patterns for all regular rooms. The results confirm the feasibility of geometry-only methods for automated, metadata-agnostic sensor visualization in digital twin environments supporting energy and indoor climate management.