Gaps at overlapping responsibilities are the likeliest place for something to go unhandled

What it is

A role-boundary gap occurs when an event spans more than one specialty or physical area and no single role owns it from onset to full resolution. This is different from a single action being claimed by nobody in the moment — that is fixed by writing down one primary and one backup for that action. A role-boundary gap is a structural gap at the organizational level: even when every role finishes its own portion correctly, the cross-specialty interface between those portions may never have been checked by anyone at all.

Why it happens

Organizations typically divide labor by equipment type or professional discipline — process, maintenance, and safety systems each run as separate lines — while faults propagate along physical connections and information dependencies that do not stop at organizational lines. Each role completes its own duty and moves on, and whether the condition at the cross-specialty interface actually holds may not appear on anyone's checklist. Role-filtered status displays make this worse: everyone sees "my part is fine," and nobody is looking at "is the interface between us fine."

There is a structural regularity tied to how many boundaries an event crosses: the more specialties or areas an event spans, the more handoff points it passes through, and even holding each individual role's error rate constant, the odds that at least one handoff was never confirmed rise with the number of handoffs — the same logic behind a series system failing if any single link fails. A coordination gap, in other words, is not evidence that some particular role was careless; it is a structural consequence of how many interfaces an event happens to cross, largely independent of who is staffing each role.

Studying it

Cross-specialty fault exercises combined with timeline tracing work well here: annotate each role's actions, handoffs, waits, and repeated steps in sequence, then map the transfers between roles to find nodes with long waits, duplicated work, or a transfer that never completed. Review should focus on the interfaces — whether both sides of a handoff actually confirmed the same condition — rather than on tallying individual mistakes, since the root cause usually is not located in any one person.

Where it stops holding

• Overlapping responsibility sometimes provides useful redundancy; collapsing every boundary into single ownership to eliminate duplication also removes the ability of one role to catch what another missed.
• Assigning an end-to-end owner for cross-domain events (a shift supervisor, for instance) narrows this gap substantially, but that role becomes a new single point of overload when several cross-domain events happen at once — closing the gap this way trades it for a different, concentrated risk that has to be weighed on its own terms.
• Ad hoc teams and genuinely novel events cannot have every possible cross-boundary case enumerated and pre-assigned in advance; some amount of gap is unavoidable there and has to be closed by rapid ad hoc coverage rather than prevented up front.
• This problem concentrates in organizations built around functional, specialized divisions of labor; in small teams where members already cover multiple roles, the boundaries themselves are blurrier and gaps show up differently.

Applying it

• Assign an end-to-end owner for cross-domain events — a role that does not replace each specialty's own work but is accountable for the event's overall closure.
• Make each role's required steps visible on a shared display, alongside the closure conditions that require joint confirmation across roles, so the interface condition itself is visible rather than implicit in everyone's private checklist.
• Route cross-domain work that no role has accepted into a shared queue with automatic escalation past a deadline, instead of leaving it under the default assumption that someone must already be handling it.
• How to check: run exercises that deliberately cross three or more specialties or areas, then audit every cross-domain handoff for a named recipient and a signoff timestamp. Handoff points that were never signed off by anyone are direct evidence of where the organizational boundary needs redrawing.

Related

• Same group: Y1.09.1 Diffusion of responsibility in operations · Y1.09.2 Coordination needs awareness of what other operators are doing, not just one's own display · Y1.09.3 Separating who acts from who checks reduces single-point omissions
• Nearby: Y7.05 Accident Investigation and Organizational Learning · Y4.05 Lockout/tagout and permits to work
• Search terms: coordination breakdown · role-boundary gap · joint cognitive system