We have become very good at collecting safety data. But how much of that data are we actually using to prevent incidents?
Having better systems, more reports and advanced technology is only part of the picture. The real challenge is turning numbers into meaningful insights and acting on them before an incident occurs.
In this edition of Safety by Qavach, I explore the importance of near-miss reporting, the limitations of traditional safety metrics, and how AI can help us move from measuring what happened to understanding what is happening.
The objective is simple: move beyond collecting data and make it count.
Read the latest edition of From Numbers to Answers.
We have become very good at collecting numbers.
Perhaps too good.
Recently, I was looking at the old 1:29:300 safety pyramid again, and it got me thinking about how much has changed in the way we manage safety– and how much has not.
We have better systems, better procedures, better technology and now AI. We can capture more than ever before. This is undeniable. But I believe we must grapple with the question- are we capturing the things that matter most, and are we doing enough with them once we have them?
There is a lot of talk about being proactive in safety. I think the real test is whether we are able to see something coming while there is still time to do something about it.
Some thoughts on that this week.
Till next time, Arjun
Do a quick search for 1 : 29 : 300 : 3,000. You’ll have something to pause over… and something to be cautious about.
This safety pyramid is likely familiar to you, and represents the relationship between incidents, near misses and unsafe acts or conditions (more here on the origins and evolution of this Accident Triangle, if you’d like to explore).
Today, we have better procedures, improved software and now automation and AI, but the shape of the pyramid remains largely unchallenged.
The most concerning bit remains the near misses. They are often under-observed, under-reported and under-actioned through CAPA. I refer to this as the “Capture Gap.”
A near miss is an early warning of a potential threat to people, machinery and the environment. If it is not captured, understood and acted upon, the risk remains. With Computer Vision– as in our ATSI– we can move beyond simply recording what happened, to identifying what is happening, enabling earlier preventive action.
For terminal leadership, this is where purpose-built QHSE software like Qavach can make a difference: be proactive and eliminate the unseen threat.
👉🏼 I wrote more on this recently on LinkedIn
I also came across this meaningful resource on near-miss reporting and accident analysis, which I believe is worth bookmarking.
Having said all that, capturing more data is only the beginning.
The real value comes from being able to see what that data is telling us.
500 inspections. 99.8% compliance. 10,000 training hours. 47 audits.
These may look like strong safety numbers. But numbers alone do not tell us what could go wrong. Most safety systems are good at explaining the past — generating reports, reviewing dashboards and showing trends.
But data itself does not prevent risk. The real value comes when data creates insight, and that insight shapes what we do next.
Much of safety data remains retrospective. The opportunity now is to move from simply measuring what happened to using data to identify risk earlier and make safety more predictive and actionable.
That is the shift I speak of– from measurement to insight, and from insight to prevention.
▶️ Watch on YouTube
I like this simple test for any safety platform: ask it a real question in plain language.
“Which part of the yard has given us the most near-misses this quarter?”
“What is my biggest unaddressed risk on the night shift?”
Most systems cannot answer directly. Not because the information is missing, but because these are system built to store data, not provide answers. You feed data in, extract it, filter it, build a report — and eventually find the answer yourself.
But a search gives you rows. An answer brings meaning. A shift supervisor, terminal manager or safety lead does not always have time to build queries and connect the dots. They have a question, and they need the answer while it is still useful.
This is one of the things we set out to change with Qavach. Ask in plain language, and get answers from your own incidents, patterns and risks.
A safety system that only stores data is a record of the past. A system you can actually ask is more like a partner in the present.
👉🏼 Read more on LinkedIn
