How to manage Line of Fire hazards with artificial intelligence?

Gökhan Yıldız
Gökhan Yıldız
-May 16, 2022

Do you know that 45% of the fatalities are due to ‘Line of Fire’ hazards? 

This was my ‘ah-ha’ moment during last year’s fatality review session at IOGP. The outcome was baffling: 6 out of 14 fatalities (45%) were classified under the ‘Line of Fire’ (LoF) category. Not long after, IOGP published the following infographics. Triggered by this insight, did I look deeper only to find out that 5 people died related to Line of Fire risks while working for Shell from 2014 to 2019 and over 40% of High Potential Incidents (HPI) were related to Line of Fire for 20202

So, it is a big killer!

What is ‘Line of Fire’?  

LoF hazards include cases where the injured person is being struck by or caught between. These hazards are not always obvious or constant and can be introduced as the task progresses (Underground and overhead lines, pipelines, objects under pressure, stored energy, lines under tension, poorly supported excavations, shifting cargo and moving equipment, etc.)

What makes compliance with the LoF rules difficult? 

The vast majority of our customers and prospects explore ways to address the increasing number of incidents related to LoF. I had the opportunity to elaborate on this topic with the major Oil & Gas (O&G) contracting companies. My key takeaways are as follows: 

  • Invisibility: The hazard is not known to the worker at the point of risks such as pressure/vacuum, velocity, electricity and tension. 
  • Shared accountability: A constantly changing work environment leads to concurrent or SIMOPS cases. For example, being hit by an object which drops from another activity taking place above with no or improper barricades. 
  • Risk normalization: Incident investigations or reviews that I have been part of reveal that the unsafe acts have been performed in the same way several times before the incident and people became complacent to the risk. As an example, deviation from a Permit is not managed so people are inclined to proceed without reassessing the situation. 
  • Lack of cross-learning: The fact that the trend of LoF incidents is plateaued informs us that organizations should learn from previous incidents more effectively. 

How can intenseye help organizations manage LoF hazards? This issue goes beyond O&G or Energy; it is a pan-industry. Intenseye offers several use-cases that can help organizations see the unseen. See a few examples that I selected for you:

Any idea, how many times do these unsafe situations occur? 

According to intenseye’s data analytics platform, these take place several times, in some cases over 100 times in a week because this has become the way how the work is performed.  

Case: How to foster a trust-based safety culture using intenseye’s real-time insights?

One of our successful customers shared these video footages with the frontline workforce and facilitated a series of engagement sessions. The leadership was pleasantly surprised that workers felt comfortable to speak up their minds and came up with solutions to make compliance easier for themselves. Our client implemented improvements that eventually decreased unsafe alerts by 80%. Additionally, they hugely benefited from smart device integration to control man/machine interactions in real-time which are the main causes of LoF related incidents. Such an inclusive approach enabled the organization to:

a) identify and fix the systemic root causes
b) secure ownership of the workers to comply
c) share these learnings with the rest of the organisation and foster trust-based safety culture

At intenseye, we believe that people at the frontline know their job best and they can offer the best solutions. We are hugely proud that intenseye’s real-time insights close the gap between organizational layers and enable the management teams to inspire those beautiful minds at the coalface to fix the systemic issues to save even more lives! 

1 IOGP: International Association of Oil and Gas Producers.

2 Based on IOGP Safety performance indicators – 2020 data. Graphs and Shell specific information were extracted from Shell’s public website:

#Blog Post
Schedule a Demo