The art of seeing

A training technique typically used in art education offers a fresh perspective on identifying hazards.

Visual literacy main
© iStock/piranka

The concept of visual literacy has been around for decades, and typically has been used to improve teaching and learning techniques in the classroom. Recently, however, it has found its way into workplaces as a skill and tool to better identify occupational hazards that could lead to unsafe incidents.

In 2017, the Campbell Institute at the National Safety Council in Illinois, US, teamed up with the Toledo Museum of Art (TMA) and the Center of Visual Expertise (COVE) for a multi-year pilot research project on how visual literacy benefits occupational safety (visit bit.ly/2RCHbDl to download the full paper). The institute enlisted four of its members – Cummins, Owens Corning, United Rentals, and USG – to examine whether visual literacy training increased awareness and recognition of workplace hazards.

Your move

An example of being unable to discern something even though you know it’s there can be demonstrated through this image. What if you knew that the chess pieces in the top half of the picture were the same colour as those in the bottom half? Can you see that?

Everything we see is based on context. We can’t see that the chess pieces are the same colour because of the different backgrounds. What happens when we make both backgrounds the same?

Now we can see that both sets of chess pieces are the same colour. Our ability to “see” things depends on visual context or distractions in the environment or background.

COVE and the Campbell Institute proposed a “train-the-trainer” model for conveying the knowledge from experts at TMA to worker populations. Each pilot organisation sends three individuals – such as environmental health and safety leads – to COVE for training. In turn, they take the knowledge back to their respective organisations to train their own team.

The question posed by the institute was simple: does “learning to see” improve a person’s ability to identify hazards in the workplace, keeping workers safer on the job? The premise is that sharper visual literacy enables individuals to perceive and comprehend more about their work environment. By enabling workers to see hazards and visualise what could occur should those hazards remain in place, they can be more proactive and intervene earlier to prevent an incident.

Understanding natural human visual biases is an important element of the skill set. There are three basic types of bias:

  • Once you see something, it is impossible to “un-see” it (see box, “Your move”).

  • Sometimes you cannot see what is in front of you, even if you know it is there (see box, “Blind spot”).

  • You are always filling in the blanks based on what you expect to be there (see box, “Fill the blanks”).

The examples of the chess pieces, leopard and the missing letters should explain our natural visual biases, and also show how our visual literacy skills can be improved when we move beyond these biases.

We are more likely to pick up anomalies and potential hazards in the workplace when we try to see the work environment (typically a place we are familiar with) through “fresh” eyes, and take care not to “fill in” visual details that we are accustomed to. Sometimes it is the fine details – and our inattention or “blindness” to them – that pose a hazard that can cause an injury or worse.

Fill in the blanks

Our minds constantly fill in “blanks” depending on what we expect to see. Take this phrase:

Can you read this?

Even when some letters are removed, it’s still possible to read:

Can yo rea thi ?

It’s still possible to read other phrases with letters missing, even if you haven’t seen the full phrase previously.

Yo are no readin thi .

Wha ar yo readin ?

Close looking

To overcome visual biases and inattentional blindness, COVE recommends a structured way of looking at work areas, which begins with taking in the big picture, or the whole scene, before focusing on individual details. The process begins by looking at the perimeter of a visual scene and then moving inwards. This process should make sense from a safety professional’s perspective because a key procedure after an incident has occurred is to assess whether an area is safe before details and evidence are gathered.

Starting by looking at the perimeter of a visual scene or work area is only the first step. Next is to look for the elements of visual literacy: line, shape, colour, texture and space. Observers should ask what potential hazards they can see by looking for indicators in these five elements: horizontal/vertical or curvy lines, shapes or angles, warm or cool colours, rough or smooth surfaces, and how much space there is between objects. This procedure and the elements/tools for looking and observing are intended to slow down the thought process so that observations can be interpreted thoroughly without jumping to conclusions.

Blind spot

Take this black and white image. Can you see the hazard? By the nature of this question, you know there is something there to see, but what is the hazard supposed to be?

When you change the image to full colour, it becomes clear what you are supposed to see.

Returning to the black and white picture, you can still see the leopard in the foreground, even though you didn’t previously. Once you see something, it is hard to remember what it was like not to see it.

Visual literacy can help in ways that go beyond the recognition of workplace hazards. As well as turning potential incidents into near-misses, having more workers trained in visual language can result in more detailed and descriptive hazard and incident reports. Improved visual language skills can be advantageous when seeking pertinent evidence from the scene. Investigators may become better at asking witnesses key questions to fill in the details of what they saw.

Being more visually literate can help safety managers to verify and audit the effectiveness of corrective actions, particularly if those actions are having the desired effect. Finally, training in visual literacy serves as an enduring safety aid, providing important learning for workers at every level. The interactivity of the exercises engages workers and leaders regularly while they receive the training that improves job safety.

To evaluate the effect of visual literacy training on hazard recognition skills, the research team is looking at both quantitative and qualitative metrics at the pilot sites. One site has been providing data since January 2018, while other sites are still implementing and delivering the training. The first quantitative metric is the number of proactive hazard recognition or near-miss reports filed. As more workers are trained in visual literacy and as their ability to see their work environments is enhanced, more potential hazards should be observed and reported. Similarly, workers may submit more orders to stop work as a result of having their eyes opened to more potential hazards.

Other metrics are related to the essential job safety analysis (JSA) component of hazard recognition. The research team suggests tracking how often JSAs are filled out and how many are completed by the full work crew. There are also qualitative metrics that are associated with JSAs, such as having a safety manager review the quality of the report. How complete is it? 

Can researchers track a convergence of vocabulary within the JSA reports as more people are trained in the use of visual language? The consistency and comprehensibility of language in the reports is important if hazards are to be managed effectively. Encouraging results have already emerged from a Cummins manufacturing site (see box below).

Pilot study results

Change in exposure, severity and fundamental risk scores for selected hazards pre- and post-visual literacy training at a Cummins manufacturing site

Falls from heightPre-visual literacyPost-visual literacy
Exposure score2228
Severity score3550
Fundamental risk score230320
Slips, trips and falls on same levelPre-visual literacyPost-visual literacy
Exposure score5074
Severity score45105
Fundamental risk score310670
Machine hazardPre-visual literacyPost-visual literacy
Exposure score242260
Severity score165180
Fundamental risk score1,2191,300

Evaluations of the train-the-trainer workshops have been positive. Based on the 55 evaluations received in late 2018, 95% would recommend the workshop to a colleague, and 100% endorsed the relevance of visual literacy to environment, health and safety.

The preliminary results from the research project and initial feedback from workers and workshop attendees seem to confirm the connection between visual literacy – primarily a skill in art education – and the field of environment, health, and safety. The Campbell Institute research team expects these initial findings to be substantiated as participant organisations continue implementing the visual literacy interventions. The team will continue to track the activities, results and lessons learned from project participants and plans to publish these in autumn 2020.

Other future directions include creating connections to other Campbell Institute research projects, such as with human performance and neuroscience, and serious injury and fatality prevention. Most of the future work on applying visual literacy will be carried out through COVE, which will develop training and support materials for workplaces.

After representatives completed the initial workshop training in Toledo, the safety team began flagging and categorising the hazards recognised by workers using the elements of visual literacy. As of May 2019, the site has expanded the training from one initial cell (work group) to other cells around the plant. Nearly 1,000 employees have been trained with 500 more to come, and visual literacy workshops are held every week. In total, 236 issues were identified using the elements of visual literacy and the incident rate decreased from 1.72 to 1.29 between 2017 and 2018.

The types of hazards identified at the Cummins manufacturing site are varied and show the diversity of areas where visual literacy can be helpful in pinpointing potential hazards. The types of hazards include slips, trips and falls on the same level, machine-related hazards, and non-safety-related hazards, which refer to hazards related to quality or production. The majority of these hazards were identified through the visual literacy elements of colour and line.

As well as tracking the number and types of hazards that employees have identified, the Cummins safety team has also compared the fundamental risk score that employees have assigned to types of hazards before any controls. Risk is assessed using the factors of severity, exposure and probability. It appears that the training has heightened workers’ risk perception and lowered their risk tolerance, resulting in higher scores for exposure, severity, and overall fundamental risk for particular hazards.


Joy Inouye is research associate at the Campbell Institute, NSC.


  • Very interesting article

    Permalink Submitted by Jon on 30 July 2019 - 07:07 pm

    Very interesting article, i have noticed that employee risk perception is on many occasions diluted by their personal experiences. This training appears to be a really good way to enhance risk perception, i believe this would be particularly beneficial to managers completing site safety audits which may not be a natural skill to them. I will look forward to seeing this kind of training in the near future.

  • I touched on this work i

    Permalink Submitted by Bridget Leathley on 13 August 2019 - 12:43 pm

    I touched on this work in O is for Observation www.ioshmagazine.com/article/o-is-for-observation so I was really pleased to see it explained in more detail here.


Add new comment