PROCESS CONTROL MONITORING AND DATA ANALYSIS

Refining, combining and otherwise manipulating materials to profitably produce end products can be a precise and demanding process. A process is a series of actions that lead to a particular result. Small changes in a process, even the minutest ones, can have a large impact on the end product. Process, as used in the manufacturing industry, refers to the methods of changing or refining raw materials to create end products. Variations in proportions, temperature, weight, viscosity, active ingredient and many other factors must be carefully and consistently monitored to produce the desired product with a minimum of raw materials and energy.  

Process Control refers simply to the methods that are used to control the variables of a process when manufacturing products. Process Control Monitoring is the tool that helps manufacturers to keep a close eye on the variables of all processes so as to detect variations early enough and correct accordingly. This is the bedrock of Quality Assurance. Everything we do boils down to monitoring processes to ensure they comply with set standards, to detect deviations and make adjustments as necessary. This is the whole point of validation or measurement (to set standards for processes), verification or comparison (to monitor deviations from standard), and calibration/qualification (to make adjustments when there are deviations). During line inspection there are data collected at set times and recorded on sheets. This is the guiding principle Of Corrective Action, Preventive Action, CAPA. You cannot correct an event or prevent it if you do not measure it. It is said that what you cannot assess/measure, you cannot understand and what you don’t understand you cannot control. What you cannot control, you definitely cannot improve.  

These datae are used to achieve many things. Inspection/review of this data and the things done with them is called Data Analysis. Data Analysis is a process of inspecting, cleaning, transforming and modeling data with the goal of discovering useful information, suggesting conclusions, and supporting decision making. 

Validation is the provision of documentary evidence that a process or system meets its predetermined specifications and quality attributes. When you do the above for Equipment it is called Qualification. Processes are validated while Equipment is qualified. You can qualify equipment without validating a process but you cannot validate a process without qualifying equipment. During validation, one must note the variables of a process. These are parameters that vary from time to time and may include weight, temperature, volume etc. 

As an example, in the manufacture of liquid oral pharmaceuticals, the process variables include weight of materials (calibration of weighing scale), dissolution and addition of materials into the mixing vortex and mixing time, compounding time after all materials have been added, filling volume in bottles (or weight for solid oral pharmaceuticals), cleanliness of bottles and water quality. If any of these variables are reduced or increased or ignored, there will be impact on the final quality of the finished product. This is why they must be monitored at all times.  

There are terminologies associated with process variables.  

1. Set Point: this is the value at which a process variable is desired to be maintained. It is the standard which the process variable must meet and remain at that level. Set points are normally presented in ranges 

2. Controlled variables: these are the variables which quantify the performance or quality of the final product. They can also be called output variables. Examples are filling volume of liquid products, filling weight of solid products, viscosity of products etc. 

3. Manipulated variables: these are input variables or factors that are adjusted dynamically or changed to keep the controlled variables at the set point. Example is filling volume or temperature or weight control knob. 

4. Error: this is the difference between the measured variable and the set point and can either be positive or negative. This is why set points are presented in ranges so as to minimize the degree of error reports which may disorganize or delay the entire process. 

All of the above are used to develop charts which are studied and used for improvement of processes. It is the sole responsibility of the QA Manager to collate all of this data, analyse them and apply them. Such data can even be used to predict maintenance activities on equipment. 

Documentation of process variables over time gives rise to lots and lots of data. This data can be analyzed by the Quality Manager and the result used for very many things. In this age of modern technology, there are machines which are capable of monitoring process variables and even generating data as evidence. No matter how efficient this machine is it still needs to be monitored by humans from time to time otherwise there will be no need to keep line inspectors in the employ of such an organization that uses the equipment. As stated earlier, monitoring process control variables generate lots of data which can be used to predict preventive maintenance. Data generated can reveal that set points fluctuate after every month. This can be used to advise the Production Manager to prepare for the servicing of the particular machine 12 times in a year since data generated from monitoring a certain process variable over a period of time indicates so. Such advice can lead to increased efficiency in production. The cost of down time in production has not been successfully quantified yet.  

Data analysis of process control variables can also be used to predict when validation of a process is due especially when errors or deviations in set points are noticed. Operators of machines can also be trained on the best way to manipulate variables and to check set points after such changes. Some operators are told to make certain adjustments every hour which they do but are not told the reason for the adjustments or the effect of the adjustments. They need to know. Instead of blaming them, they need to be encouraged to report deviations from set points so there can be adjustments and documentation otherwise data can become distorted if deviations are concealed.