The objectives of this project were to:
-Develop a dynamic scheduling system for the factory based on Knowledge-Based System (KBS) techniques. The system was to bridge the time gap between the large planning horizons of logistic systems, eg COPICS/MRP II, and the real-time conditions on the sh op floor (microplanning).
-Develop generic shells for the following: knowledge acquisition modules both for factory analysis and factory operation; user interface modules for different kinds of operators, including management in the factory using KBS techniques and multi-windowin g; and expert systems for production planning, preventative maintenance and quality control.
-Build these modules, implement them in two very different plants (electronic appliances and tyre manufacturing), and test the applicability of one generic shell to the different applications.
The programme included the development of:
-software modules for plant data acquisition, real-time plant data updating, interpretation KBS, and plant simulation
-KBS for maintenance, quality and process planning
-knowledge-acquisition modules for GRAI and SADT analysis methodologies.
Tyre quality assistant (TQA) is a knowledge based system (KBS) for the intelligent support of a quality manager (QM). The system considers about 60 defects and 300 corrective actions in the production of industrial vehicle tyres. The knowledge base has about 500rules, about 20 database static objects (with hundreds of slots), and several tens of database dynamic objects (actions). The system is currently installed in an Italian tyre factory of the Pirelli Group.
Innovative aspects include:
integration of conventional software technology and knowledge based systems (KBS) technology;
integration of tyre quality assistant (TQA) with the already existing factory information system;
an intelligent decision support system (IDSS).

The objectives of this project were: firstly, to develop a dynamic scheduling system for the factory based on knowledge based system (KBS) techniques. The system bridged the time gap between the large planning horizons of logistic systems, and the real time conditions on the shop floor. Secondly, to develop generic shells for the following: knowledge acquisition modules both for factory analysis and factory operation; user interface modules for different kinds of operators, including management in the factory using KBS techniques and multiwindowing; and expert systems for production planning, preventative maintenance and quality control. Thirdly, to build these modules, implement them in 2 very different plants (electronic appliances and tyre manufacturing), and test the applicability of one generic shell to the different applications. The programme included the development of: software modules for plant data acquisition, real time plant data updating, interpretation KBS, and plant simulation; KBS for maintenance, quality and process planning; and knowledge acquisition modules for analysis methodologies. The decision network acquisition tool (DNAT) has been programmed and the decision network analysis of the factories completed. The controller for the surface mounted devices workcell has been designed. The general system design has been finished. The diagnosis KBS has been programmed. Demonstrations are available for a workcell controller, artificial intelligence (AI) simulation, a diagnosis KBS, a planning KBS, a commonality analysis of electronic products, a planning KBS, a DNAT tool, and a knowledge acquisition tool.
The decision network acquisition tool (DNAT) has been programmed and the decision network analysis of the factories completed. The controller for the SMD (surface-mounted devices) workcell has been designed at Philips. The general system design has been finished for the dynamic planning task at Pirelli and BICC. The diagnosis KBS has been programmed for Pirelli. Demonstrations are available for:
-workcell controller at PFH
-AI simulation of FMS at PFH
-diagnosis KBS at ARS
-planning KBS at BICC
-commonality analysis of electronic products at PZI
-planning KBS at AEG
-DNAT tool at Graphael/SGN
-knowledge-acquisition tool at Politecnico di Milano.
The use of KBS for dynamic scheduling will enable a rapid response to changes in requirements and in machine and material availability. The first KBS was introduced into the Pirelli factory during 1987. Piloting of workcell controllers in Philips CIM production line took place in late 1988 for short-term production planning, and was followed by quality and maintenance controllers in late 1989. The final result of the project was the description of the various KBSs developed by each partner following the project's CIM-AI Implementation Guide.