Human Error In Manufacturing Is A Systems Problem

When something goes wrong on the line, the first instinct is often to blame the operator. But most of the time, the real problem isn’t people, it’s the system.

Manufacturers today are running tighter margins, shorter cycles, and more complex operations. Under that kind of pressure, even skilled operators make mistakes and especially when instructions are vague, processes are fragile, or the equipment setup doesn’t match the task.

Human error is part of any operation. But in most cases, it’s a signal that the system needs attention.

This guide looks at those underlying issues and shows how leading manufacturers reduce mistakes by improving system design instead of assigning fault.

Manufacturing has a human error problem. Or does it?

A growing body of research in psychology and organizational studies has changed how we think about workplace mistakes. We now know that most individual errors happen when systems make human error likely, or even inevitable.

In other words, human error isn’t actually a human problem. It’s a problem with work systems, along with the lack of flexibility to augment worker processes.

In this post, I’ll explain why, and show that understanding the reasons humans make mistakes is the key to error-proofing your operation.

[Download Our Guide] Poka Yoke: The Ultimate Guide to Error Proofing Your Processes →

Recent research by the DOE defines human performance with a simple formula.

Human performance is the consequence of behaviors and results. Or, Performance = Behavior + Results.

In this definition behaviors are things people do and say. They’re observable acts that can be seen, heard, and measured, encompassing both the mental and physical efforts required to accomplish a task.

Results are the outcomes of behaviors.

Simply put, behaviors are a means, results are an end, and performance is a measure of the consequences of a specific combination of the two.

The key finding of the DOE report is that behaviors and results are influenced by a variety of factors beyond the control of front-line workers.

The authors found that:

“Human error… is not a cause of failure, alone, but rather the effect or symptom of deeper trouble in the system. Human error is not random; it is systematically connected to features of people’s tools, the tasks they perform, and the operating environment in which they work.”

This definition accepts as an axiom that humans will make mistakes. To err, after all, is human.

But it holds that errors that result from human nature–defined as mental strain, biases/assumptions, limited attention resources, attitude, limited perspective, and fatigue, among others–are predictable, and therefore preventable.

Because we understand exactly what aspects of human nature and behavior will precipitate errors, it is possible to institute the controls necessary to stop them.

The report argues that many errors are the result of “latent organizational weaknesses,” or flaws in project design, management, or company culture that riddle an operation with “error traps.”

Error traps can be many things. They can be an inherent part of a task, such as a manual assembly too variable or complex to be performed consistently. They can be the result of poor management or training. Or they can be less tangible, like bad attitudes or a work environment that glorifies overworking to the point of fatigue.

Whatever the specific cause, there are some commonalities shared by all “error traps.”

• High stress and time pressure exacerbate poor human performance
• Many tasks implicitly ask humans to exceed the limitations of human capability, inevitably leading to mistakes
• Error is more likely within complex systems
• Workers are unreliable narrators of their own ability to maintain control in adverse working conditions

The report lists 5 principles of human performance. Together, they help explain why errors occur and suggest methods for preventing them.

1. Even the best people make mistakes – Accepting that people make mistakes is the first step toward constructing systems that minimize the potential for error. It’s important to ask what elements of your operations could be contributing to suboptimal performance. As the report states, “Understanding how and why unsafe acts occur is the essential first step in effective error management.”
2. Error-likely situations are manageable – As the saying goes, “knowing is only half the battle.” It’s critical to proactively “minimize the presence of conditions that provoke error” once they’ve been identified. “Event-free performance is… dependent on ensuring the integrity of controls, controls, barriers, and safeguards against the residual errors that still occur,” the report notes. In manufacturing, digital technologies can help.
3. Don’t overestimate the importance of culture – If “management is the business of directing workers’ behaviors,” building a culture that respects and avoids risk is central to minimizing error.
4. Workers respond to encouragement and reinforcement – Cultural change doesn’t happen overnight. Positive change happens when workers begin to assimilate a new technology or culture and are recognized for doing so. “Because behavior is influenced by the consequences of a worker’s experience, what happens to workers when they exhibit certain behaviors is an important factor in improving human performance.”
5. Mistakes will happen. Progress follows when mistakes become lessons – Iterating is the key to achieving value fast. While many manufacturers already apply this agile thinking to their continuous improvement initiatives, it can also be fruitfully applied to moments of failure.

When a frontline operator makes a mistake, it’s rarely about attitude or effort. More often, the system itself makes the correct action harder to take than the wrong one.

Many recurring design issues lead to the same kinds of errors:

Whenever a process depends too much on memory, multitasking, or personal experience, errors multiply. When tools and systems carry more of that load i.e. by showing, confirming, and adjusting in real time, operators can focus on doing the work instead of catching up with it.

It’s clear that humans aren’t leaving manufacturing anytime soon. The question thus becomes, How can manufacturers design work systems sensitive to the limits of human nature? How can manufacturing systems work with people to maximize human performance?

Given what research has shown about human error, helping people perform can require many different actions. Some are small, like encouraging healthy habits in and outside the workplace or using new technologies to decrease workers’ cognitive load. Others, like changing attitudes throughout an organization, are more challenging and require broad alignment on goals and actions.

Additionally, augmentative technologies like digital work instruction, IoT connected inline quality checks, and no-code development platforms can improve working systems and decrease errors to close to zero.

The DOE, however, summarized the reasons to take action best when they wrote,

“No matter how efficiently equipment functions; how good the training, supervision, and procedures; and how well the best worker, engineer, or manager performs his or her duties, people cannot perform better than the organization supporting them.”