Closed-loop manufacturing, a STEP-NC process for data feedback: a case study

摘要

This article exposes a solution to ensure Closed-Loop Manufacturing from CNC machines to CAM systems. The main goal is to capitalize knowledge from the CNC machine and to reuse it in the CAM system. This proposal aims to help CAM programmers in making choices for planning new machining sequences. The STEP-NC standard allows bi-directional exchanges in the digital chain - from CAD systems to CNC machines. Unfortunately, STEP-NC standard does not allow having a traceability of machining program. To achieve the information feedback from CNC machine to CAM system, Closed-Loop Manufacturing sets up a manufacturing loop using PLM systems. This solution is based on OntoSTEP-NC, an ontology based on STEP-NC which provides trades between Product Data Management, Manufacturing Process Management and Enterprise Resource Planning. As a solution OntoSTEP-NC supports data extraction from CNC machines, and relevant information reinjection to CAM systems. Closed-Loop Manufacturing is a three steps process: (1) Manufacturing feature recognition to have the most relevant information reinjection from the Manufacturing Process Management database in the CAM programming stage, (2) validation of machining sequences before capitalizing them, (3) capitalization of cutting parameters, the manufacturing feature and the material to fill the Manufacturing Process Management database. This process will help CAM programmers making choices for cutting conditions based on company good practices. A case study has been chosen to test Closed-Loop Manufacturing concepts. The case study is divided into three scenarios aiming at testing first the "reuse phase" including manufacturing feature recognition for the reinjection of relevant information. The third scenario tests the "capitalization phase" and the ability of the system to detect the changes made by the operator directly on the CNC machine. This paper also details the trades between the CAM systems and OntoSTEP-NC by focusing on extracted data and the way the information is capitalized into the Manufacturing Process Management database. It also details the queries addressed to the Manufacturing Process Management database to send the most relevant information.