Shewhart recognized two types of variation, namely the special-cause variation’ and the ‘common cause variation, he’s the inventor of control charts. The Shewhart model shows two types of variation can be called ‘assignable cause’ and ‘chance cause’, respectively. He created a control chart to help explain these two types of variations. Shewhart suggested new variables and attributes to his control charts. Shewhart suggested that common-cause variation be controlled in order to improve quality and decrease scrap. This will allow any process to be controlled statistically. To distinguish between common and special causes, it is necessary to first bring the process under statistical control. It is possible to predict future outputs and manage processes economically by bringing a process into this state. Shewhart’s principle opened the door to the modern scientific analysis of control processes.

His skills were developed while working for Bell Telephone, where he was responsible for improving the reliability of the transmission systems and later the voice clarity of carbon transmitters in company’s telephone handsets.

Shewhart model

Shewhart’s research focused on reducing variation within a manufacturing process. He defined the problem as assignable-cause or special cause variation and created the control chart to help distinguish between them.

Any industry should strive to create economic methods that satisfy human needs. This can be achieved by simplifying things so that they require very little human effort. It was possible to establish economic limits on routine efforts by using scientific techniques and modern statistical theories. Any routine that results in deviations from these limits is considered to be economically unsustainable. To make the process economically viable, it is necessary to identify and eliminate the problem.

Walter Shewhart was modest, but he was an innovator. He discovered a way to improve process consistency and measure it. This may seem mundane, but it was important back in Bell Labs when it was all the rage. He was focused on the idea of predictable processes in order to make consistent money for the company.

His shewhart model focuses on improving the voice quality via phone lines. He decided to separate the results from the call by asking “Can you hear me now?”

  • Predictable Some things happen on calls all of the time. Low voices were difficult to hear. The longer the distance, the less quality or thickness of wires could make a difference. This was “common stuff”.
  • Unpredictable Some things don’t happen very often. Heavy ice from a snowstorm could cut off voices completely, or lines too close to bridges might make a call static. We’ll call this “special stuff”.

He concluded that a process can remove, prevent, or mitigate “unpredictable stuff”, then it can be controlled. A predictable and controlled process is the only way to meet customer expectations – “yes, I can hear you now!”