March 3, 2012
Product Performance vs. Process Control
Remember that early class or seminar you took on Statistical Process Control (SPC)? You know, the one that said if you can learn to understand and control the critical inputs of your process, the critical outputs would take care of themselves. This is a great concept, except for one thing: my customers aren’t interested in the speed I run my line or in the temperature of my drying ovens when the product passes through - the critical inputs.
They are more interested the length of the product or the hardness of the product after it has been cured - the critical outputs - because these are the characteristics they use once they purchase my products to assign specification limits. On the other hand, my manager, while caring about the critical outputs, wants to understand the critical inputs which affect profitability and allow us to manufacture products more efficiently and lower our operating costs. Ultimately, both groups are looking for the best product, for the lowest price. Let’s see how the Data Management System (DMS) and the Data Collection Service (DCS) can help us accomplish both goals and allow us to create meaningful relationships with our data.
The first step in our data collection process is to understand where our raw data is stored. Since our data is normally stored in multiple locations, we don’t collect data together because these systems do not talk with each other. In some cases, I might be able to get one system to talk to another, but I may not have the structure to create a meaningful result. These systems might include: an ERP system, such as Oracle or SAP; an MES system; or a process control system that uses Programmable Logic Controllers (PLC’s) with an OPC server or run through an application such as Wonderware. Taking it a step further, in addition to these systems, our output data may come from various gages or only be available in a text file.
The next step involves setting up the various providers of DMS to collect all of the needed information. This may seem daunting, but DMS gives us the tools to grab the required information from each of the systems mentioned above. We need to set up the OLEDB provider for the data from the ERP or MES system, the OPC provider for the data from our PLCs, the Wonderware Live Provider for the data from our Wonderware application, and the Grid Provider for the data from our gages and text files. Each provider is set up independently to grab its specific data at our defined intervals so that the data is available to create subgroup data in our ProFicient system using DCS.
DCS allows us to collect the information that DMS has amassed and allows us to define how often the data is collected and under what conditions it is collected. Once we define how often the data is collected, we can define the descriptive elements of our data set, such as part, process, lot number, date, time and any other user-defined elements needed for the data set.
Once these are identified, we define the process results we need (line speed, belt speed, oven temperature), along with the product results (length, thickness, and hardness).
DCS saves the data to our ProFicient database as a unified subgroup of both process and product data. This allows us to chart all the information we’ve collected based on the product we’re making and the process used to make the product over time. DMS/DCS has given us the ability to provide our customers with the product data they need, while giving our company the bigger picture view of how the process settings relate to the finished product results. With this knowledge we can improve our manufacturing efficiency and increase profits.