OPINION: A New Approach to Maritime Safety is Needed

For years, the shipping industry has focused on regulations and procedures to improve safety. Yet shipping is still at risk of major accidents. The whole industry needs to change its focus. Ticking boxes never made anyone safer. Also, assessing culture using valid and reliable survey instruments can help to improve safety.

It has been well documented that most maritime accidents (~80%) are caused by human error. Still, most of the focus on learning is rooted in technical causes and adding procedures and checklists.

Despite this bias, many accident investigation reports pinpoint that the leadership or safety culture was the root cause of more recent accidents such as the Bulk Jupiter, El Faro, Helge Ingstad and Costa Concordia, as well as older accidents such as the Exxon Valdez, Bow Mariner, Herald of Free Enterprise and Amoco Cadiz.

Industry blind spot
The critical failures leading to the accident were in most cases known before the accident took place. This demonstrates that failures which are not handled properly may develop into critical situations and accidents. This is a blind spot because the biased focus on technicalities and “impeccable” safety inspections makes people reluctant to be open about their failures, concerns and mistakes.

We at SAYFR think shipping companies, and the whole industry, needs to change its focus. Thousands of auditors and inspectors across the world are engaged by classification societies, flag and port state authorities, vetting and insurance companies and HSEQ departments. They verify that ships do the right thing and comply with technical and procedural requirements. However, ticking boxes never made anyone safer.

Cover-up culture
Also, and worryingly, there is a cover-up culture causing errors and unsafe practices. There are now so many procedures and checklists that, in some cases, it is impossible to comply with all of them. The fear of failure is driving accident statistics, and surveys reveal that 45% of seafarers admit that they regularly do not comply with procedures. 

I firmly believe that human factors are key to prevent threats and failures from escalating. Yet improving safety or performance is about improving not only individuals but also the collaboration between sea and shore staff, between officers and crew and between different nationalities and cultures on board ships.

Huge potential to reduce accidents
Although this is recognized, it is not always addressed, so I believe a new approach is necessary to improve collaboration and reduce risks. Indeed, collaboration is strongly correlated with the risk of accidents and business interruption. Our experience of working on multiple projects over the years shows that it is possible to reduce the risk of major accidents by up to 75%.

However, there is no quick fix to improve collaboration and implement behavioral changes through, for example, training courses. Changing the culture is key and that process takes time. To help operators improve their approach to safety, proven methodologies must be used.

Safety leadership behaviors
Put simply, it involves observing and identifying working methods on board and then working with all the officers and crew in teams and as individuals to deliver the eight-point safety leadership behaviors, namely:

• Giving feedback
• Speaking up
• Building trust
• Creating openness
• Showing care
• Facilitate learning
• Promoting teamwork
• Managing dilemmas

Experience shows that the focus on the eight behaviors work because they address the blind spot. By encouraging the participants to openly share errors, failures and concerns, they are able to break the chain of events that can lead up to a major accident. Also, this approach helps to move beyond the culture of punishment to the positive safety-enhancing culture where crew members help each other.

Culture assessments key to improving safety
In order to understand how the organization culture influences safety, there is a need to use methodologies specialized for this purpose. One thing that many people are ignorant of is that a key professional competence of organizational psychology is advanced mathematics and data analysis. The evaluation of organizational culture relies on interviews, observations and questionnaires applying psychometric instruments that are tailor-made to ensure valid and reliable results. The professionals drive the process while the data provides the results. As a consequence, the more and better the data on these topics, the more valid, reliable and to-the-point are the results.

However not all the instruments used in the industry are valid and reliable. In a recent review of safety culture maturity instruments, only 3 of 43 instruments were valid. Indeed, there is not one single test alone that can demonstrate the validity of a survey instrument. Therefore, SAYFR has developed tailor-made psychometric instruments to assess these topics and has a database of responses from about 300 000 seafarers.

When it comes to the qualitative and quantitative tests that can be made to verify validity, the basic one is content validity. This dictates how well a safety instrument addresses a safety issue. It specifies that the survey instrument adequately covers the topic being studied as well as having sound scientific grounds and references.

This is important because so many historical examples show risks that were identified well ahead of time but were not addressed. These include the Deepwater Horizon blowout, which claimed 11 lives and caused huge environmental damage, rig personnel had knowingly by-passed safety barriers. In this case, failures were identified but the root cause of the problem – i.e. human neglect, whether cultural or circumstantial – was not factored in.

Reliability of survey instruments through data
Reliability of the survey instrument is also key and that is ensured statistically by use of data. Factor-analysis is a statistical method used to describe variability among correlated items in terms of a potentially lower number of unobserved variables, called factors. For our instrument, the eight factors are equal to the eight SAYFR leadership behaviours (8SLBs) mentioned above.

Moreover, predictive validity is the instrument’s ability to predict something in the future such as an event, or correlations with instrument measurements made by other instruments. If an organization scores low in terms of the 8SLBs, it is a good indicator of future problems. This has been shown on a number of occasions when departments, units or suborganizations have received low 8SLB scores only to have accidents occur in the intervening time, before action was taken.

Predictive validity can also be applied to solutions. When action was taken based upon a low 8SLBs score, a shipping company experienced a 60% reduction in the frequency of serious accidents, to a level which was maintained five years subsequent to the investigation.
Concurrent validity and construct validity are also important elements. Concurrent validity measures the correlation between two independent measurements performed at the same time. An increase in the ability to manage failures, for example, will necessarily correlate with the number of incidents that occur.

Construct validity is when a theoretical model of cause and effect – for example, do the improvements prescribed following appraisal of the 8SLBs – accurately replicate the real-world scenarios they are intended to represent? Construct validity is the ultimate validity measurement, and necessarily incorporates all other validity factors.

Reduction in the frequency of serious accidents
Also, it’s not only the psychometric instruments that rely on data. The use of digitalization, the internet of things (IoT), sensor data, machine learning, and big data has picked up in recent years. The idea is that those with the most data can create the best analytics and forecasts. With the use of more quality data, risk assessments and worst-case scenario simulations provide reliable predictions and identify effective interventions to prevent accidents.

In short, what we at SAYFR see is that the best shipowners and operators have a proactive organizational culture that goes beyond ticking the ‘compliance boxes’ and instead applies a collaborative, trusting approach from top to bottom in the company’s organization. This also includes assessing culture using valid and reliable survey instruments. This is what really helps to improve safety.

About the Author: 
Dr Torkel Soma is a senior partner at SAYFR. He manages the company’s research activities and core safety insights. He holds a doctor degree in using artificial intelligence and machine learning methodologies to measure safety performance in organizations