Improving Quality Control with SPC

Statistical Process Control (SPC) was initially developed over 65 years ago by Walter A. Shewhart as a method to improve or achieve quality control in a manufacturing process. In this method, statistical tools like frequency distribution, histograms, scatter diagrams and Pareto charts are used to view and reduce process variation, thereby reducing defects and waste from the production line. Although SPC is a cost-effective and thorough method of determining and predicting significant deviations in a production environment, quality control managers routinely face inconsistent results when implementing this method manually. Human error is a major source of inaccuracy along various points in the process. For instance, misinterpretation of the display readout, incorrect transcription onto printed documents, and even weighing the wrong sample can skew overall statistical results.

This is consistently overlooked in manufacturers that utilize manual SPC in their production lines, although it can lead to lost profitability and increased manufacturing costs in the long run. In order to avoid these costly issues, companies must move to higher technology methods for collecting, analyzing, and storing SPC data, thus reducing and even eliminating the human error factor.

Hidden Problems of the Current Manual SPC Process

To better understand the core problem and apply an efficient, effective solution, it is pivotal to understand how quality control managers in the food industry utilize manual SPC. A recent visit to a food packaging plant highlighted inconsistent results of SPC on the production floor. Generally, 15 to 20 samples are tested from their production line at different times during the day. These samples are then individually weighed, and a production worker manually records the results on a clipboard. The individual weight results are then entered manually into a spreadsheet program, and various statistical calculations are performed. The results of the calculations and charts are then used to adjust the filling machines in manufacturing to deliver a consistent result. If necessary, the quality control manager must resolve any conflict that is revealed through the results. Conflicts can arise from normal wear and tear of the machinery used on the production line, or even from fluctuations in initial raw materials. If they are not corrected, the product container can be over-filled, which causes lost revenue, or under-filled, which violates government and regulatory guidelines.

The results from this food packaging plant produced no predictable over-filling or under-filling statistical trends despite the fact that the same people used the same scales at the same time of day. Normally, sample weights will fluctuate slightly over time due to the manufacturing and filling process, and individual operator behavior. Operator behavior can be accidental or deliberate. For example, if one is hurried, numbers may be transposed on the hand-written documents, or the same sample may be weighed multiple times. On the other hand, if an operator thinks that they will be reprimanded for irregular weight readings, or they don’t want to stop the production line for adjustments on the hot filling machine, they may be tempted to enter false data on the hand-written documents.

Per Figure 1, there are many steps in the manual SPC method that can result in problems with the finished product. In addition, this method offers no traceability of weighing results throughout the process. Therefore, the underlying result of the manual SPC method is a loss of profit to the company when each package of food product is either over-filled beyond the acceptable weight range or under-filled, resulting in hefty government penalties and fines. The implementation of an automated SPC scale system would eliminate these manual user errors with only a few changes, providing peace of mind and long-term cost savings.

Automated SPC System Advantages

The fundamental purpose of an automatic SPC system is to remove the manual element from the employee’s responsibility, thus greatly reducing or eliminating human error associated with recording incorrect sample weight results. To accomplish this, all manual weighing apparatus and documentation methods are replaced by a scale system equipped with automated SPC software. This software is fully configurable to satisfy all SPC tolerances.