|
Critical Parameter activity comes in two distinct forms: first is the application during technology and product development and is aptly referred to as Critical Parameter Development; second is the application of the knowledge developed to manage critical parameters during production operations including supply chain management. Both PRODUCT and PROCESS critical parameter relationships must be identified and characterized by the development team.
Managing critical parameters during production and supply chain operations is much less complicated than developing the detailed knowledge that proves certain parameters (Xs and the Ys they control) are indeed truly critical. As we have stated in previous newsletters, there are 7 checks that can be applied during technology and product development to objectively prove that Ys as a function of Xs are exhibiting behaviors that have gone past being classified as high risk and are now proven to be objectively problematic. The Big 7 metrics for establishing critical status include: Measurability, Stability, Adjustability, Independence / Interactivity, Hyper-sensitivity, Robustness and Capability. If you have Xs and Ys that fail any one or more of these metrics - then they are deemed critical. This is true for both PRODUCT and PROCESS control and functional response parameters.
As the product development team concludes their work, they should publish a "Design Guide" that contains all the knowledge that is needed to manage the product and its assembly and manufacturing processes during its post-launch life cycle. Probably the most important chapter is the one focusing on the Ys and their controlling Xs that have been deemed critical. This sends a warning to the Life Cycle Management (LCM) team that they have special actions and instructions to safely measure and control the critical parameters.
The special actions and their control methods and best practices include:
Measurability Problems: In general, the LCM team will need to establish an ongoing set of measurement quality assurance protocols, based on the advice from the Product Development team, to assure all measurement systems used to gather sample data on critical parameters are under control and capable of providing trustworthy data. This will include ongoing evaluation and continuous improvement of measurement Bias, Stability, Linearity and Repeatability/Reproduce-ability Studies.
Stability, Adjustability, Interactivity, Hyper-sensitivity & Robustness Problems: Certain Ys, functional response variables, that have instability problems must be placed under an appropriate continuous variable control charting plan. The same must be done for controlling Xs that have a history of instability as they are used to control manufacturinfg PROCESSES or part/material characteristics that are manufactured and applied to contribute to the control of the PRODUCT'S functions.
Control Charting is the best practice used to spot patterns of shifting and drifting away from statistical, stable control. The LCM team must apply statistical PROCESS control on the machines and processes that make the materials and parts that integrate into supporting the functions of the product. They must recognize the difference between SPC (Statistical PROCESS Control) methods for identifying the first onset of assignable cause variation as it attacks manufacturing and assembly PROCESSES vs. Statistical PRODUCT Control methods that identify within part/material characteristic variation as well as part-to-part and batch-to-batch variation.
They must conduct Correlation Studies to link PROCESS variation effects to PRODUCT variation effects. They must have objective data to prove they understand the transmission of variation from PROCESS to PRODUCT functional performance. This transmission of variation must be classified as "Critical-to-Function". This designation tells everyone in the LCM and Development teams that this transmission path directly affects specific functions in the PRODUCT (by way of the parts and materials) that cannot be left to chance. These relationships must be placed under a rigorous inspection process using appropriate Control Charting methods to measure impending failure and to provide ample warning to adjust or stop the production PROCESS before it makes out-of-spec. materials or parts.
The PROCESS must be adjustable to put it back under control. It's too late to only inspect parts and materials to manage CPs. It is common for PROCESS critical parameters to be placed under 100% inspection using within-process metrics to assure no materials or parts are allowed to be made. This is pure preventive Process measurement and control. PRODUCT inspection in this context is redundant and should only be conducted in extreme cases where human life and safety is at stake. If Correlation Study data is rigorously documented for these cases, you know a priori that when the PROCESS is adjusted before bad parts or materials can possibly be made - you are practicing CPM!
If PRODUCT and PROCESS development professionals have done their CP tasks diligently, then the LCM team already knows a great deal about what to measure and when to measure it in the PROCESSES. They already know where the hyper-sensitivities and instabilities exist and what to watch for. They also know what the adjustment parameters are to control these problematic parameters.
Capability Problems: Once PROCESS and PRODUCT control and correlation is established as mentioned above, capability studies can be conducted as frequently as is economically and ethically required. Cp and Cpk indices will clearly show where the CPs stand with regard to requirements in light of measured variation and mean performance.
My advice is biased strongly towards rigorous management of CPs in the production PROCESSES once development is complete. This is the cheapest and most appropriate time and place to do this work - which I view as problem prevention at the root cause. This is about managing problems, not reducing risk. We know which Ys and Xs are proven to have significant problems relative to the Big 7 metrics. If we focus on proactive applications of Measurement System Analysis, Control Charting, Adjustment back-to-target schemes and Capability Studies, we are doing the best practices that are the state-of-the-art in modern critical parameter management.
| 
