Quality is a strategic consideration for any manufacturer, and is a driver of both profits and customer loyalty. While many people know that quality is a high priority concern, there are some conventional practices that limit progress.
Bridge the Silos
Many organizations see the limits of information “silos” in either manufacturing or technology as primary constraints. These separations occur for a variety of reasons. New investments in management systems may be completed incrementally, addressing specific areas of the operation separately, automation systems and new production equipment my not be configured to exchange information easily or data exchange may be a secondary consideration when evaluating system design.
Regardless of why silos happen, we must always think a few moves ahead, making plans for integrated data sharing across all related systems. Process control systems, machine control systems and environmental control systems are all operational layers that you must consider, as each of them will impact the quality of your products. By connecting your data sources, you will able to see a holistic view of your operation.
Leverage the Historian
As you deploy and run your automation systems, your repository of process settings and measurements will grow as they are logged over time. Most commonly, this information is used to diagnose and troubleshoot problems after they occur. These historical trends, however, can also be used to understand your plant’s performance across a long time period — leverage this information for additional insights to improve overall performance!
Analysis of this historical information allows understanding of the best and worst cases for machine and process setups. You should be able to identify root causes for operations that are less than optimal and look for new opportunities to increase quality.
Analyzing Every Step of Your Process
Quality is a cumulative result. Take a broad view of your operations and model the complete production system: from raw materials that enter your operation, to processing and production operations, to finished goods. By taking note of every step of your process, you will increase total system visibility and bring insights to light that may have been missed when systems were not connected. Isolated systems are literally blind to upcoming problems, reducing response and increasing operational variance.
How can you effectively monitor your system as a whole? Gather data early in the process to understand performance of raw materials by considering the effects that environmental factors may have on your materials. Then, gather midstream data to understand your production and process performance, understanding how your various processes interact to achieve your final output. Data later in your process will share information related to the performance of your finished product and final packaging operations.
In each area, review the measurements you are taking and consider how well you are able to model the overall operation to understand your entire system’s performance. Is your model static or dynamic, with real-time updates?
Beyond Real Time: Predicting Outcomes
Your plant’s performance is partly due to process design and partly due to how well it operates. While real time information is helpful in monitoring operations and able to advise you when things go wrong, it’s similar to using historical data to tackle problems after they occur. By trending your performance over time and building models that predict where your processes are going, you will be able to better predict when problems will occur and prescribe corrective actions earlier to avoid losses or defects and maintain quality.
If you are able to recognize silos and information gaps in your plant, you have the opportunity to fill in holes and improve the performance of your existing operations. By including a predictive forecast for better and faster outcomes, you’ll take the quality of your products to the next level.
Is quality your first priority, or is something else a bigger issue? Let us know in the comments section below.
Estimated Reading Time: 3 minutes