today, the real waste isn't the error. It's discovering it too late

giugno 03, 2026

When it comes to quality, the goal is often expressed very simply: eliminate errors. It's a phrase everyone agrees with. No one wants to produce defective components. No one wants to manage complaints. No one wants to deal with rework or scrap. Yet, observing many industrial processes, I've become convinced that the main problem isn't the error itself. The real problem is when that error is discovered too late.

Errors are part of any process.

It may seem provocative, but no industrial process is immune to errors. Even in the most structured organizations, there are variables that are difficult to completely control. A misinterpreted quota. An out-of-date revision. A feature not transferred correctly. An unexpected movement. An actual condition different from that assumed in the design phase. Thinking of completely eliminating every possible error is often unrealistic. It's much more realistic to build systems capable of rapidly identifying anomalies before they propagate throughout the process.

An error discovered immediately has a limited impact. An error discovered after additional work costs more. An error discovered after heat treatment costs even more. An error discovered during assembly has further consequences. An error discovered by the customer often represents the worst-case scenario. The problem isn't just financial. The further an error progresses in the process, the greater the number of activities involved, the time lost, the necessary checks, and the number of people who must intervene. This is why the true cost is almost never the initial error. It's the delay in identifying it.

Why prototypes exist

One of the most common misunderstandings concerns the role of the prototype. Many see the prototype as an additional cost. In reality, its primary purpose is very different. A prototype is used to uncover problems when they are still economical to correct. It is used to test hypotheses. It is used to validate design choices. It is used to understand where actual behavior deviates from expected behavior. In other words, it helps to advance the discovery of errors. And this is precisely its value.
The same reasoning applies to PPAP, dimensional checks, and traceability. They are often perceived as bureaucratic activities. In reality, their original objective is very concrete. Reduce the risk of realizing too late that something isn't working. Of course, documentation alone doesn't guarantee quality. But if used correctly, it can help expose problems that would otherwise emerge much later.

The Poka-Yoke approach teaches the same lesson.

The Poka-Yoke concept follows the same logic. It wasn't created to identify errors. It was created to prevent them from spreading. The goal isn't to ask people to be perfect. The goal is to design systems that make it difficult to make mistakes and easy to immediately identify an anomaly. Because an error caught immediately is manageable. One discovered weeks later can become extremely costly.

Very often, the most costly errors aren't caused by technological limitations. They arise from a lack of visibility. Information distributed across multiple systems. Unaligned reviews. Processes that depend on individual experience. Checks that arrive too late. Incomplete communications. In these cases, the problem isn't the component. It's the time that passes before someone realizes something isn't working.
Speed, too, matters in quality
In recent years, there's been much talk about speed as a competitive factor. We often think of development speed or speed to market. But there's another speed that's becoming increasingly important: the speed with which an organization can identify and correct its errors. The most effective companies aren't necessarily those that make the fewest mistakes. Very often, they're those that learn the fastest.

The real question

Perhaps the most useful question today isn't: "How can we eliminate every error?"
But rather: "How long does it take us to notice an error when it occurs?" Because that's precisely where a key part of industrial efficiency comes into play.
Errors will continue to exist in any organization. What really makes the difference is the ability to intercept them when they're still small, cost-effective, and manageable. Prototypes, controls, documentation, traceability, and robust processes all have a common goal: not to create bureaucracy. To reduce the gap between the moment a problem arises and the moment it's discovered. Because today, the real waste isn't the error. It's too late to find out.

BLOG CATEGORIES

Blog

MECHANICS IN PILLS

LAST NEWS FROM MECHANICS