Semantic Models And State Tracking For Plug & Produce In Automation Systems

Abstract

Modern manufacturing systems must be highly flexible, scalable and dynamic to adapt to changes necessary to deal with mass customization issues arising as result of changing customer behavior. The production systems must have the capability to integrate and remove system components in a short-time and ideally on the fly to meet changing product and customer requirements. This can be achieved by developing a hierarchical production system composed of dynamically configurable system components capable of interacting with each other. This thesis presents a methodology to develop configurable information models for shopfloor devices, state machine models for tracking the states of system building blocks and their interaction mechanism.

This research work aims to explore information models and standards being used in Process and Factory Automation industry. It involves semantic modelling of production resources and exposing their functionality to an information modelling and machine-to-machine (M2M) communication tool using Service Oriented Architecture (SOA). It presents a mechanism to interlink information models for bi-directional flow of information across the layers of manufacturing system. It also proposes a methodology to build a scalable hierarchical manufacturing system with state-tracking capability of systems and sub-systems on each hierarchical level. The proposed solution allows reusability of information models which makes system deployment and scaling easier and faster. Finally, the developed concept was tested on a manufacturing system and state-tracking, interaction between system components and bi-directional event-propagation were achieved.