One Safety Picture: Aligning Risk Across Interfaces Without Losing Operational Meaning

You departments aren’t necessarily seeking blanket standardisation. They’re trying to work with an aligned meaning of Safety risk.

You’re not dealing with a “risk matrix problem”.

You’re dealing with a meaning alignment problem.

Across any transport organisation—aviation, rail, bus, maritime, different departments rarely set out to impose blanket standardisation across interfaces where safety risk analysis is concerned. What you see instead is something more operationally honest (and more dangerous if left unmanaged):

each department uses risk language in good faith
each makes locally sensible judgements
each believes its scoring is reasonable in context
and yet, when leadership asks for a single risk picture, those interpretations simply don’t line up

Not because people are incompetent. Not because they are trying to control each other. More so because, risk is contextual, and departmental interfaces amplify the differences.

So the question isn’t “How do we force one matrix?”

The question is:

How do we align interpretations of likelihood, severity, and tolerability across interfaces—without stripping away operational context?

That is the real work of governance.

Why “One Safety Picture” breaks at the departmental interfaces

A single safety picture fails when it tries to merge risk ratings without reconciling the the liklihood and severity assumptions underneath.

At interfaces, departments often differ on:

what the risk statement is actually about
(hazard presence vs loss of control vs outcome)
what likelihood means
(exposure frequency, probability of failure, probability of consequence)
what severity represents
(credible worst case, expected outcome, bounded outcome)
what controls are assumed to be in place
(and who actually owns them)
how quickly conditions can degrade
(tempo, workload, degraded modes, competence, supervision)

So when executives ask for a single view, teams aren’t resisting alignment. They’re struggling to translate meaning.

If you compress those differences into a common colour or label, you don’t get alignment.
You get false equivalence, and false equivalence is exactly how latent risk hides inside safety reports.

Key insight: “Tolerable” is not universal—
tolerance is contextual

“Tolerable” sounds like a category you can standardise. In practice, it behaves like a boundary condition that shifts with context.

Most departments are not trying to redefine tolerability for political reasons. They’re reacting to different realities:

Different exposure profiles
A risk that is “rare” in one department may be “routine” in another purely because of frequency of contact.
Different recovery opportunities
Some contexts allow detection and correction before harm. Others don’t.
Different control strategies
One area relies on engineered barriers; another relies on procedural or human performance controls. These can degrade differently.
Different time horizons
Projects and change programmes often accept temporary risk with compensating controls; frontline operations often cannot.
Different consequence bearers
The same top event may land consequences on a different group than the one scoring the risk.

So tolerability becomes difficult to align because departments are not only rating “risk”—they are implicitly rating:

control confidence
degradation sensitivity
operational margin
ability to monitor and recover

This is why “One Safety Picture” cannot be built by averaging ratings, converting colours, or issuing a single template.

A credible single picture requires a method to align meaning at the interface.

What good looks like:
harmonisation through interface governance

You still need a single risk picture for leadership decision making, assurance planning, and safety performance oversight.

But you build it by harmonising interpretation, not by forcing uniform scoring.

That means:

using aviation SMS likelihood and severity terms with implicit definitions
requiring operational context to be stated
(conditions, phase, exposure, degraded modes)
treating interfaces as potential safety risks, not admin boundaries
explicitly reconciling tolerability differences where risk is shared or handed over

In other words: the enterprise picture is not a collage of departmental heatmaps.
It’s a governance view of interface reconciled risk.

The missing capability:
a Risk Interface Alignment Sheet

Most organisations don’t lack matrices. They lack a practical alignment mechanism.

A Risk Interface Alignment Sheet is a simple tool that helps two (or more) departments align:

what is being assessed
what assumptions are being used
what controls are depended upon
what tolerability means in shared environment
what must be escalated and assured

Used properly, it becomes the bridge between departmental risk analyses and the enterprise safety picture—without stripping away operational context.

If you want the sheet and guidance on how to apply it across your interfaces, we’ll walk through it with your real examples. Get in touch with us, to arrange a meeting.

What changes when you fix interface alignment

When you treat interface alignment as a governance discipline:

the risk translation exercise becomes easier
disagreements become visible early, not during or after incidents
escalation becomes faster and calmer
assurance targets control effectiveness where it matters
leaders gain a single safety picture they can trust under pressure

And most importantly: your “One Safety Picture” becomes operationally credible, because it respects how work is actually done right across the organisation.

Call to action

If you’re trying to align interpretations of risk across departments and you want a practical method (not another template), email me at contact.us@aviaintelligence.com to book a discussion.

Not ready for the next step? No problem, register with our Zenler community at the
The Safety Rebels Club