Statistical Process Control (SPC) is the use of statistical analysis and techniques to manage a process. Generally, the purpose of SPC is to make processes related to production more efficient, using monitoring numbers to eliminate wasteful actions. While it is inherently geared toward manufacturing processes, SPC can be applied to any operation that relies on a process of actions to reach an output.
“Control”, as it is used in this context, is a loose term that calculates the managing party’s ability to consistently produce a specified outcome from each process. It infers accountability--as in, which people or automated systems are in control of each process--but also effectiveness. The statistical methods used in SPC are to provide empirical data that managers use to foster continual improvement of processes.
New methods of process control are created all the time, but these are a few of the most popular.
Control charts are diagrams that measure whether the individual processes in a given system are under control; specifically, how much they vary from a perceived norm or intended mark. A control chart consists of a graph representing optimal or normal output for a process, which represents ‘control’, or the constant in the scientific method. Other measurables, which are sometimes overlaid to show variation with the control graph, will express the highs and lows of actual output during different segments of a process. The variation of the output might deviate within expected parameters and cause no concern, or it might vary widely enough to call for an examination of the process. It is up to the managing parties to decide which variations are excessive and which are stable.
A check sheet is a detailed listing for collecting process data, generally a table of events or complications measured with figurative check marks. Measurements in this case can refer to any number of definitions, so long as instances can be recorded with ‘check marks’ in a straightforward manner. For example, a truck driving company might use a check sheet to quantify the obstacles of a problematic route. The sheet might be a list of any incident that slows down the drivers route such as high-traffic streets or road construction, where the goal is to use the check sheet to see if any changes in process can expedite the trip. The company might decide to do away with the route altogether if the process variation isn’t manageable.
Many other tools are employed to instill SPC, and each involve some form of statistical math for tracking variation, as well as a way to measure which parts of the process are affecting which. In a control chart and some others like it, process stability metrics are used to inform decisions about which processes require immediate attention, if any. The criteria for “stability” is as flexible as the criteria that establishes “control”. Stability might mean that a process is producing within expected highs and lows; or the expectation might be higher, at which point it might still be considered “out of control”.
SPC is often associated with System Quality Control(SQC), but they are not synonymous. SQC measures the quality of outputs, whereas SPC measures the effectiveness by which these outputs are achieved. Of course, measuring outputs informs the effectiveness of processes, and efficiency cannot be implemented at the cost of quality, so the two methods often work together.
The main benefit of using SPC, compared to other methods like SQC, is that it is preemptive. SPC is geared toward meeting challenges as they occur instead of correcting outcomes after they’ve happened, which translates into business as minimized risk compared to a reactive method.
Beyond improving their quality, SPC also eliminates the waste in processes, eliminating redundancy and making each of them more efficient. Quality improvements ensure that businesses are putting out the best products or services; waste removal increases the speed of these processes. Supplying a better product more efficiently is the best way to edge out competition, particularly in manufacturing.