Quality in a Manufacturing System

The literature on Quality does not dwell on its interactions with other components of a manufacturing system, like Production, Engineering, Production Control/Logistics, or Supply Chains. As a consequence, it is missing out on key relationships that affect the value of quality improvement.

Quantity versus Quality: A Conflict?

In particular, the literature presents quality and quantity as separate pursuits, missing out on their synergy. In the field, the relationship is often conflictual. This goes back to the time the first inspector rejected a unit that a producer wanted to be passed. As recounted by David Hounshell (1984), the concept of tolerances, now taken for granted, emerged in the 19th century specifically to mitigate this conflict.

In the 20th century, quality became a profession, a support department in manufacturing organizations, with a professional society, a specialized body of knowledge, and certifications, separate from the design of production lines or production control.

This separation is artificial if you consider manufacturing from a systems perspective. You don’t just need high quality because the market demands it but also because you cannot produce in volume without it. Ramping up production without a capable process is driving on a road full of potholes.

Deming’s View of Manufacturing as a System

When advocating the appreciation of manufacturing as a system, Deming drew a high-level map of materials and information flows:

Another Perspective Manufacturing as a System

Deming did not make quality a component of the system. The following diagram shows some of the system interactions between Quality and other functions within manufacturing. It’s about the work of maintaining and enhancing quality, not about who does it, and it’s admittedly not exhaustive. Highlighted in red are reinforcement loops.

How quality interacts with other components in a manufacturing system

The work of quality is shown as including four subjects:

  • Capability. The ability of processes to meet expectations in terms of critical product characteristics is the foundation. Establishing, sustaining, and enhancing it has been the focus of SPC 100 years ago, Taguchi DOE 40 years ago, and applications of data science and AI today.
  • Response. Even capable processes are subject to occasional, discrete breakdowns that cause the Quality department to receive or issue problem reports. The response ranges from filling out a form documenting the solution to recalling products.
  • Compliance. Government agencies and customers issue mandates that the company must comply with to stay in business.
  • Mistake-proofing. Capable processes with rapid problem response are still vulnerable to human error, which is prevented by mistake-proofing their operator interfaces.

It is not a complete list, as it does not include the calibration of equipment, training, the organization of recalls, or statements to the public when disasters strike. This incomplete diagram is intended to highlight a few relationships:

  • The most vital in most manufacturing organizations is the way efforts on quality nurture the quality reputation of the company among its customers. In many markets, it is its crown jewel, but it is not the only interaction that matters.
  • Once you have capable processes, you can organize production in flow lines and pull between lines, which allows you to detect and respond to quality problems an order of magnitude faster than when processing batches in a job shop.
  • As the organization solves problems so that they do not recur, they become rare, and this lets human error percolate to the top among the remaining defect causes, which you address by mistake-proofing the processes.

Economics of Quality

Even recent American texts, like Pyzdek and Keller’s (2013), justify efforts in quality through “cost savings” that are aggregations of claims made over many projects about costs of failure, appraisal, and repair. They ignore the interactions highlighted above.

While not necessarily easy to quantify upfront, these interactions are the reasons to pursue quality improvement beyond what narrowly defined costs of quality would justify. They also explain the sequencing of the approaches:

  1. First, you establish process capability so that you can implement flow and pull.
  2. Then you leverage rapid problem detection to solve problems.
  3. Once problems are sufficiently rare, you switch your focus to mistake-proofing.


  • Hounshell, D. (1984). From the American System to Mass Production, 1800-1932: The Development of Manufacturing Technology in the United States. United Kingdom: Johns Hopkins University Press.
  • Baudin, Michel. (2001). When to Use Statistics, One-piece Flow, Or Mistake-Proofing to Improve Quality.
  • Pyzdek, T., Keller, P. A. (2013). The Handbook for Quality Management, Second Edition: A Complete Guide to Operational Excellence. United Kingdom: McGraw-Hill Education.

#quality, #systemsthinking, #manufacturing, #costofquality