Knowledge-Based Service-Oriented System for the Automated Programming of Robot-Based Inspection Systems

Abstract

The automation of inspection tasks using robot-based inspection systems (RISs) is a complex procedure that is partly still done manually and requires expert knowledge from different fields such as systems-engineering, robotics, and metrology. We propose the use of a knowledge-based system as an elementary step to model and generalize the required knowledge to achieve the full automation of RISs. Its core component, a knowledge base for RISs, is introduced in this paper. Using a three-layer modeling approach the required expert knowledge is formalized and categorized into three different types: descriptive, procedural and relational. The knowledge base is implemented using ontologies to provide a hardware-neutral, reusable, and modular system model. Furthermore, we introduce a service-oriented architecture that uses models from the knowledge base to define the communication interfaces to ensure the data integrity and exchange between the knowledge base and all software modules of an RIS (e.g., robot and sensor control, user-interface, and simulation software). The validation of the proposed approach is shown by using the example of an industrial RIS.