Fusion Of Multi-modal Data With Semantic Model Towards Providing Situational Awareness

Abstract

We propose the concept of situational awareness in a complex industrial environment which provides the operators with decision making capabilities based on information extracted from heterogeneous data sources ranging from structured databases to IoT based sensor data and semi-structured, unstructured reports. We implement this concept using a semantic model (e.g. a Knowledge Graph) which allows traceability and relationship discovery across entities through graph navigation and inference. Our solution consists of a framework of multiple authoring services to create a unified industry specific knowledge graph from the multi-modal data sources.