Team cognition in remote maritime operations: a scoping review and cross-domain theoretical synthesis

TomAlright, here's something on my mind: Team cognition in remote maritime operations: a scoping review and cross-domain theoretical synthesis. Where do you stand on this?

CarolYou are asking me where I stand when we still don't even agree on what "team cognition" looks like on a bridge at three in the morning. It’s like trying to diagnose a relationship problem without knowing if the couple is actually talking to each other or just sitting in separate rooms. The review says our concepts are fragmented, which basically admits no one knows how these crews are actually coordinating under pressure right now.

TomThree in the morning isn't when coordination breaks down; it's usually been drifting since port. A crew that can't track shared intent during routine shifts will shatter under stress because they're relying on heroics instead of a working rhythm. That fragmentation you mentioned means we're building safety cases on guesswork rather than observed behavior.

CarolDrifting since port assumes the baseline is stable, but the data shows even routine checks vary wildly between vessels because nobody standardized what counts as effective communication yet. We’re arguing about whether the ship sinks slowly or fast while the crew is still figuring out how to pass a message without dropping critical info. Until we define the mechanics clearly, every safety model is just educated speculation wearing a lab coat.

TomCalling it speculation ignores that sailors have been passing messages across decks for centuries; the issue is that the old tricks don’t translate to satellite links where half the context gets lost in transit. Without a clear map of how those signals change, you can’t tell if a delay is bad weather or poor teamwork. That gap is what keeps the whole field spinning its wheels.

CarolSpinning wheels misses the human cost; when that context evaporates over a sat link, a junior officer might freeze instead of escalating because he has no reliable cue that his partner heard him. That hesitation is where real errors get born, not in some abstract model. We need to look at who specifically freezes up when the channel goes quiet, not just theorize why the theory is messy.

TomFreezing is a symptom, not the cause. The junior officer hesitates because the team never established a clear signal for “acknowledged” versus “heard,” leaving him to guess if his partner is actually engaged. You can’t fix the fear by mapping the anxiety; you have to design the feedback loop that removes the uncertainty in the first place.

CarolDesigning that feedback loop sounds clean until you realize most crews improvise their own shorthand, so a rigid protocol clashes with the habits they’ve already formed at sea. You can’t mandate a universal “acknowledged” signal without testing how it lands on a tired deck hand who’s been operating on instinct for twenty years. The real danger is forcing a sterile system onto people who survive by adapting to chaos.

TomYou’re describing a protocol war, but the actual problem is silence. If the deck hand’s improvisation works when the radio crackles, mandating a sterile code doesn’t help anyone hear through the static. We need systems that amplify their existing cues, not replace them with bureaucracy that fails the moment things go sideways.

CarolAmplifying existing cues assumes everyone is sending the same signal, but a veteran’s nod means something completely different from a rookie’s grunt over a noisy channel. If you build the system around the loudest habit, the quieter operators become invisible precisely when you need their input the most. That bias toward the experienced crew is what leaves the whole watch vulnerable to blind spots.

TomMaking the quiet ones invisible is exactly what happens when you judge a shift by volume alone; the rookie’s grunt carries the same weight as the vet’s nod once you stop listening for authority and start listening for confirmation. On a remote watch, silence from the junior is louder than any shout from the senior.

CarolListening for confirmation isn't a strategy, it's a hope; on a distributed watch, there's no shared room to verify who actually understood the message before acting. You're asking juniors to trust their gut in a vacuum where the seniors are physically isolated from their immediate consequences. That reliance on intuition is a liability, not a virtue, unless the technology explicitly bridges the physical distance.

TomTrusting your gut is fine until the gut says everything went well while the sensors say otherwise; bridging the distance requires explicit data feeds, not just faith in the operator. A junior officer needs a hard indicator that the command was received and executed, not a vague sense that the ship is still moving forward.

CarolHard indicators sound like the solution until you account for the lag; by the time that green light pops up on the console, the maneuver may have already gone wrong downrange. You’re trading one kind of blind spot for another—replacing real-time judgment with delayed digital paperwork that makes the operator feel safer than they actually are.

TomLag kills the illusion of safety, but zero information guarantees disaster. The answer isn’t choosing between raw instinct and delayed screens; it’s fusing both into a single view where the delay itself becomes part of the alert. Without that fusion, the operator is just guessing with extra steps.

CarolFusing the two sounds elegant, but it ignores who has to pay attention to both at once; a fatigued operator staring at a screen trying to reconcile a sensor ping with their own instincts is doing double work under pressure. The system shouldn't demand higher vigilance from human beings; it should handle the reconciliation automatically so the operator only sees what matters right now.

TomAutomating the reconciliation creates a new failure mode: if the algorithm gets the weighting wrong, the operator trusts the lie without ever seeing the conflicting data. Until we can prove which source—the sensor or the gut—is more reliable under fatigue, you’re just swapping human error for machine error. The review never answers how we validate that automatic filter, so we’re stuck designing for uncertainty instead of eliminating it.