|
PROCEED Webinar Series
Data Mining in Manufacturing: Optimizing Product Quality
A STATISTICA Webinar in the PROCEED Webinar Series
May 8, 2006
Data Mining in Manufacturing
Incredible advances in the data mining of large data repositories have arisen in recent years driven primarily by the needs in a variety of industries such as Retail, Insurance, and Financial Services. In the manufacturing industry, we have witnessed the rise of Six Sigma and other data driven approaches. Manufacturing companies are now sitting on a wealth of data from the last few decades of investment in data collection and storage.
Until the development of the PROCEED software, two critical factors have delayed the application of data mining techniques to process improvement and product quality optimization in manufacturing:
1) Injecting Data Mining Expertise into the Optimization Process
DMAIC - Improve Phase:
The ability of process and product subject matter experts to inject their expertise into the optimization process to specify the constraints of equipment tolerances, fixed parameters related to availability of raw materials, and the costs to implement recommended process changes. Before PROCEED, data mining methods outputted an "answer" that may not have been feasible to implement. With PROCEED, subject matter experts interact with the application during optimization to perform what-if scenarios that align with what is feasible and practical to implement.
2) Gaining acceptance of solutions across the plant and enterprise
DMAIC - Control Phase:
With data mining techniques, the other critical item is how to get buy-in across the plant or enterprise that the solution is reasonable with those stakeholders critical to its implementation. In addition, how do you deploy the predictive model to those stakeholders in the process that must use and understand it day to day to control the improved process? The answer is PROCEED's "Actionable Decision Environment."
Register to view the recorded webinars.
Caterpillar Uses the PROCEEDTM Software to Simultaneously Improve Quality and Throughput
A STATISTICA Webinar in the PROCEED Webinar Series
October 14, 2005
During this webinar, Caterpillar and StatSoft reviewed a case study in which Caterpillar used the PROCEED software to improve the product quality and streamline product testing to meet customer demand.
The PROCEED software is a turnkey manufacturing software solution that distills fundamental causal relationships between your products and the processes that produce them, using the data that you already collect and manage. The PROCEED software implements the patented approach developed and proven at Caterpillar Inc. and powered by the STATISTICA EnterpriseTM Analytics Software Platform.
The PROCEED software combines novel and traditional knowledge extraction methods to:
- Derive and validate simple to complex causal relationships between manufacturing processes and product quality outcomes
- Deploy actionable information that enables process owners and knowledge workers to compare what-if scenarios and simultaneously optimize multiple competing outcomes
PROCEED's Actionable Decision EnvironmentTM is the mechanism for deploying knowledge into action. Process owners and decision makers are empowered to manipulate process settings to observe the impact on outcomes. For example, setting a particular process parameter at its theoretical optimum may require a costly upgrade to the production process not justified by the benefits of the change or may be beyond the capabilities of the supplier of the materials. PROCEED provides that flexibility for process owners to interact with the model to inject their subject matter expertise to derive the "real" optimal settings.
Register to view the recorded webinars.
|
|
