Ray Harkins — Active Contributor

Why “Good Enough” Assumptions Often Are

Co-authored with Mike Vella

If you’ve spent any time working with real manufacturing, reliability, or field data, you already know an uncomfortable truth:

Most statistical models assume ideal conditions that rarely exist in practice.

Textbooks often begin with assumptions like perfectly normal distributions, clean random samples, and well-behaved processes. Meanwhile, engineers are dealing with skewed cycle times, mixed populations, censored failure data, process shifts, and the occasional mystery outlier that refuses to explain itself.

Turning Uncertainty into Informed Action

Co-authored by Mike Vella

At its core, risk management exists to protect us. It aims to reduce both the likelihood and the consequences of adverse events, whether those events affect safety, quality, cost, equipment, or project schedules.

Good risk management is not reactive. It is proactive. It assumes that future events can be influenced and, to some degree, controlled. That belief alone fundamentally changes how organizations plan, design, and operate.

Attribute inspection is one of the most widespread and difficult-to-control measurement methods in manufacturing. Whether inspecting machined surfaces for cosmetic defects, checking weld quality, reviewing molded parts, evaluating assembly completeness, or using go/no-go gauges, many operations depend on human inspectors to make subjective judgments. [Read more…]

In conversations about “Six Sigma level” performance, I hear people using DPPM and DPMO like they’re the same thing. And in fact, they’re close, but not interchangeable. Among manufacturing quality professionals though, this distinction matters critically when describing how good (or how bad) a process really is.

(Originally published in Quality Magazine, October 2025)

The term Measurement Systems Analysis (MSA) refers to a collection of experimental and statistical methods designed to evaluate the error introduced by a measurement system, and the resulting usefulness of that system for a particular application. In manufacturing, measurement system quality directly affects decisions about processes, products, and a manufacturer’s ability to meet their customer’s requirements.

In the push toward digital transformation, automated inspection, and AI-enabled quality systems, it’s easy to overlook the fundamentals that quietly hold everything together. Among them are two terms that are frequently confused, even by seasoned professionals: Calibration and Verification

These words live at the core of measurement assurance, and understanding the difference isn’t academic; it shapes risk decisions, audit outcomes, and ultimately, product quality and reliability.

In manufacturing, variation is the enemy of quality. Left unchecked, it erodes reliability, increases waste, and drives up costs. One of the simplest but most effective tools for controlling variation is Layered Process Auditing (LPA).

Of the many constructive soft skills, active listening is surely one of the most helpful in the design thinking process. More than just listening to collect information, active listening is a mindset that fosters empathy for the speaker. In business and engineering, it’s our customer that’s speaking. Active listening seeks to understand not only the other person’s words, but also their attitudes and motivations.

For many people, even the idea of “being a better listener” seems ludicrous. Once you’re listening, how can you listen better? It’s like swallowing better or breathing better. [But the truth is, you can.] Even the best listeners periodically fall into a handful of traps that obstruct their understanding of their employer’s, customer’s, and loved one’s words.

Early in my quality management career, while working at a small extrusion and fabrication company, I learned something important: bosses pay attention to the money. And if I focused on cost savings projects, I could stay on their good side.

Most of my cost savings efforts at that time focused on eliminating specific types of defects. After all, even a low-frequency defect—especially one that reaches a customer—can drive substantial savings once resolved. Other projects looked inward, targeting inefficiencies in our systems and practices. Lab procedures, control plans, and audit schedules tend to drift out of sync with the products and processes they’re supposed to control. So every now and then, a little system hygiene—an organized cleanup—can free up resources and allow you to reallocate attention to where it’s needed most.

It was during one of those hygiene projects that I stumbled into something I’ve since come to call The Paradox of Invisible Discipline.

A Practical SPC Method for High-Capability Processes

By Ray Harkins and Michael J. Vella

Manufacturing professionals who regularly apply Statistical Process Control (SPC) methods, such as X̄ and R charts, know the power of SPC to maintain control of both quality and cost. The standard practice emphasizes attention to variation, with operators trained to take corrective action whenever a process signals an out-of-control condition. Typically, this means adjusting a process input, verifying the correction, and resuming production.

However, not every out-of-control signal requires intervention, especially in processes where inherent variation is extremely small and product quality is consistently high. In some cases, such as tooling wear or small material variations, process center shifts may occur outside of the operator’s ability to correct. Responding to these shifts with standard SPC rules can lead to overcorrection, lost productivity, and increased variation.

In manufacturing and quality assurance, we rarely have the time—or budget—to inspect every single part in a batch. That’s where acceptance sampling comes in. It’s a statistical tool that helps us decide whether to accept or reject a lot of material based on a random sample. But more than just a shortcut, acceptance sampling is a structured decision-making process that balances quality expectations with efficiency and risk.

If you work in manufacturing, quality, or engineering, you’ve probably heard the term “Design of Experiments,” or DOE. Maybe you’ve even helped collect data for one without fully understanding what was going on behind the scenes. That’s okay. DOE is a sophisticated statistical method, but even if you’re not the one crunching the numbers, understanding the big ideas behind it can make you a more effective technician, engineer, or manager.

At its core, Design of Experiments is a structured statistical approach to testing. It helps us understand how different inputs (or factors) affect an outcome (or response). Whether you’re adjusting a process to improve tensile strength, dialing in machine settings to reduce defects, or figuring out which supplier provides the most consistent material, DOE helps answer one big question: What settings or conditions will give me the result I want?

Brainstorming may sound like a casual conversation technique, but when applied properly, it becomes a critical tool for continuous improvement across a range of creative and technical disciplines.
In a structured environment, brainstorming enables organizations to capture a wide range of ideas, perspectives, and solutions — ideas that might never surface in a traditional problem-solving meeting.

Let’s take a step back and revisit the basics of effective brainstorming, especially its role in driving meaningful change.

In the world of quality, reliability, product design, and manufacturing, improvement is a necessity, not a luxury. Few models have provided a stronger foundation for improvement than the Deming Cycle, commonly referred to as PDCA: Plan-Do-Check-Act.

Although simple in structure, PDCA represents a deep and disciplined approach to learning, problem-solving, and continuous improvement. Whether you are optimizing a production process, refining a laboratory method, or developing a new product, understanding PDCA is essential.

Continuous improvement doesn’t happen in a vacuum. To raise the quality, reliability, and performance of their products and processes, organizations must look beyond their own four walls.
Benchmarking — the practice of studying others to improve your own performance — is a foundational tool for competitive advantage.

Whether comparing products, processes, or management systems, effective benchmarking helps organizations discover new ideas, set realistic goals, and accelerate improvement.

Let’s take a step back and explore the basics of benchmarking: what it is, how it works, and why it matters.