Pertinence, a software vendor whose solution allows companies to improve their industrialization processes, today announces that Metaltemple, designer and manufacturer of aluminum parts for engines, gearboxes and other automotive components and a subsidiary of Teksid Aluminum, the world leader in the production of aluminum castings, is to deploy Pertinence Rule Maker to improve the manufacturing process of some specific parts.
Metaltemple designs and manufactures parts based on client specifications, in a high-volume production environment. In order to guarantee production quality while simultaneously reducing costs, the company is continually on the lookout for ways to improve its processes. These processes are controlled by various parameters (metal temperature and composition, casting parameters, etc.). The impact of these parameters on the quality of the parts can vary significantly. At the end of 2002, the company began working with Pertinence Data Intelligence to improve the manufacturing processes for certain parts, for which the rejection rate varied between 15 and 20 percent. This high rejection rate resulted in significant financial losses.
Metaltemple mapped out the complete production process to structure data collection. Pertinence Rule Maker was then used to generate rules from this data. These rules were used to prioritize the parameters that could impact the quality of the production process. The solution uses automatic learning techniques to identify the critical production parameters from a total of fifty involved in the overall process.
Pertinence Rule Maker has allowed the company to identify three influential production parameters and to generate associated rules from manually collected data. These rules help Metaltemple's experts understand the complex interactions governing their production methods.
"Pertinence Rule Maker is easy to use and does not require any advanced statistical skills, but relies instead on an excellent understanding of the manufacturing process, explain Bonnet and Fayolle Technical Directors at Metaltemple. "We are impressed with the user-friendliness and ease with which the system was implemented. We have benefited from considerable timesavings in both the development of our processes and their implementation. The correlations are quickly highlighted, the rules are comprehensible and easy to interpret, and the analysis is easier and much more precise. Moreover, the rejection rate has considerably de-creased, to about five percent."
The system is currently installed on three workstations. The new version, Pertinence Rule Maker 2.0, will facilitate integration with the Manufacturing Execution System (MES) by July 2004. Working in conjunction with the MES, Pertinence Rule Maker will allow them to periodically verify production rules. As a result, Metaltemple will be equipped to significantly reduce process ramp-up costs, and to accelerate the industrialization process.
"We have saved about 300,000 euros since implementing Pertinence Rule Maker and have already recovered our initial investment." Bonnet and Fayolle conclude by saying, "Besides, we have seen tangible results in about a year, a relatively short time. In addition to the financial benefits, we have a much more reliable manufacturing process, which is very important, given the more than 30 million parts manufactured each year. Pertinence Data Intelligence helps us oversee complex manufacturing processes in a clear and simple way, in order to optimize our production activity and rise to the challenge of continuous improvement."