Semantic and Structural Interoperability in Environmental IoT A Unified Conceptual Framework for Industrial Odor Surveillance

Sensors are deployed and streams are transmitted, yet inference on odorous episodes frequently fails to improve, because the data remain mutually unreadable even when they are mutually transmissible. This article argues that the persistent difficulty is not a sensing problem but an interoperability problem with two inseparable faces : a structural face, concerned with how heterogeneous observations are placed on a common temporal and representational substrate, and a semantic face, concerned with whether the aligned observations can be read in a shared vocabulary.

We propose a unified conceptual framework that treats each observation as a typed entity encoding its time, space, uncertainty, and provenance, and each transformation as a contract-bound morphism declaring the distortion, latency, and validity terms it imposes. The framework organizes industrial odor surveillance around four archetypal streams source, emission, vector, and impact maintained at their native rhythms, and around four architectural layers mediating between physical capture and decision.