Share this article:

The misconception of human error

André Campos CO-CEO Tangível

There is nothing better than a story to prove a point. And what a story we have here! On 28 May 1979, at the Three Mile Island nuclear power plant in Pennsylvania, U.S.A., a partial nuclear meltdown almost resulted in a catastrophe similar to Chernobyl. It was the worst nuclear-related accident in the United States to date.

At the root of the accident were failures in a non-nuclear secondary system, which were intensified by a malfunctioning valve, resulting in the release of major quantities of nuclear coolant.

The twist of this story (since a good story always has one) is that the operators made no attempts to close the valve, a fact that left everyone puzzled in the following days, including President Jimmy Carter, who visited the site.

During the investigation of the accident, it was found that the valve had a visual indicator in the control panel that was providing incorrect information. Contrary to assumptions, this indicator did not show whether the valve was open or closed, but instead whether or not the valve was being supplied with power.

This is why the operators did not attempt to close the valve during the accident, which would have been enough to stop it from happening. Therefore, the design of a single button almost caused an extremely serious nuclear accident.

According to Donald Norman, the world’s leading expert in usability and UX, “the control room and computer interface at this nuclear power plant could not have been designed in a more confusing manner, even if they tried”.

This is a clear example of a system design that is poorly executed, resulting in mistakes that could then be attributed to people, while ignoring the true reason behind the failure: a design flaw.

We constantly hear, see and read this in the media: accidents, contingencies, problems caused by human error. Erroneously, I would say! Many of these incidents have often covered up a key aspect of the system designer’s responsibility.

These days, regardless of our profession, we are surrounded by systems that have been designed by someone. There is still very little care taken in what is known as “Human-Centered Design”. As a result of this lack of care, from a highly superficial standpoint, mistakes are attributed to people who then feel bad because they do not know their way around the system when, in fact, these mistakes are caused by the system’s designer.

To avoid these mistakes, it’s important to involve professionals specializing in usability or human-centered design when creating most systems (especially the digital) that we use each day.

If the systems that everyone uses to operate nowadays in the digital world are not designed for people, bearing in mind those who will operate them, we are guaranteed to have lots of problems. On the one hand, if the system is not well-designed, the learning curve of its operators will inevitably be long, leading to frustration and low productivity.

Furthermore, if the information of a given field is not explained well, it can result, due to a lack of understanding, in mistakes leading to low productivity or much more serious situations, namely economic loss or even risk to human life (such as a nuclear accident).

These failures originate at the time of designing the system. An app, website or any other product must be conceived with the involvement of designers who know how to make systems for people.

This involves having professionals who are properly trained, together with skills originating from social sciences such as anthropology, psychology or sociology (in order to truly identify the needs, frustrations, and anxieties of people), as well as designers with specific training in usability and ergonomics, among others.

No system should be launched without the appropriate usability tests. A usability test consists of having people from the target audience of the product or service use the system under conditions very close to those of the real world, observing how easy it is for them to properly use the product or service.

Often, this is a process that unfolds as the system matures, beginning in a very early design phase, which can help us understand why many of the available paths are not the most appropriate.

When the product is launched, just like functional tests are done from a technical standpoint, it’s essential to perform usability tests. In this way, when there is a propensity for certain user errors, there is still time to correct these flaws. As the old saying goes, better safe than sorry.