Research Papers on navigation with the use of Visual Observations

Professional Mariner Forum
1

Eye tracking during high speed navigation at
sea: Field trial in search of navigational gaze
behaviour

High speed craft. From the paper:

Results: The results show that the novice drivers
look at objects that are close to themselves, like instrumentation, while the experienced look more at objects far away from the boat. This is in accordance with previous research on car drivers. Further, novice boat drivers used the electronic navigational aids to a larger extent than the experienced, especially during high speed conditions. The experienced drivers focused much of their attention on objects outside the boat

Scan Pattern for the Maritime Navigator

image
image266Ă—205 6.01 KB

Rotterdam Pilots spent 90% of the time looking out the window.

image
image264Ă—158 6.05 KB

image

I’ve been using the term “visual navigation” which implies the instruments are not being use which is not case, “visuals observations” is a better term. Instruments are used but for a much shorter "dwell time’ because information from instruments typically can be read at a glance.

5 Likes
2

I use the term “piloting” for the same practice of using both visual navigation and instruments.

We test our officers on it using the bridge simulator at Seattle Maritime Academy. A six-hour long simulation Coastal Transportation’s Training Group developed, focused on Discovery Passage, Seymour Narrows, etc.

4 Likes
3

These papers are written by human factors (HF) researchers, they use the term “navigation” in the broad sense including keeping the vessel in safe waters and collision avoidance.

“Visual observation” is used only to mean out the window , while “scan” or “scan pattern” is used for both outside and the instruments.

I’ve skimmed a half dozen or so papers, apparently they don’t use the terms “piloting” or “visual navigation”

Eye tracking
AOI: Area of interest
AOIO Area of intrest outside
Dwell time.

The linked paper Scan Pattern for the Maritime Navigator uses a lot of acronyms but defines them in the text.

1 Like
4

Very good article:

It’s worth reading the from the beginning but here’s some key sections:

Visual observation forms the foundation of lookout technique, yet atmospheric conditions frequently restrict what eyes can see.

Excessive information availability creates paradoxical performance degradation. Seafarers spend cognitive resources tracking information updates rather than analyzing situation significance. Attention fragments across multiple displays. Decision quality decreases despite information quantity increasing. The gap between information supply and processing demand widens.

The transformation occurs when seafarers recognize that RADAR-based lookout functions differently from visual observation. Visual lookout provides continuous real-time awareness—peripheral vision catches movement, color indicates ship type, and lighting patterns reveal vessel status. RADAR presents discrete updates at fixed intervals, abstract representations requiring interpretation, and delayed information that may not reflect instantaneous reality.

Experienced seafarers develop filtering strategies:
â–ş Identify priority information for immediate attention
â–ş Ignore non-critical data temporarily
â–ş Process high-importance items thoroughly before moving to secondary concerns
â–ş Revisit filtered information periodically to catch changes

This selective attention maintains performance despite information overload. Inexperienced seafarers attempt processing everything simultaneously, becoming overwhelmed and ineffective. They collect RADAR data on all visible targets, plot every ship on screen, sequentially review ARPA information, yet fail to identify which vessel requires immediate attention.

Ships already passing safely at 5+ miles rarely require ARPA tracking. Visual or RADAR observation confirms their safe passage without consuming acquisition capacity.

This is the key point, direct visual observation is more efficient the using the screen.

The cognitive workload of interpreting RADAR displays, distinguishing targets from clutter, and extracting movement information simply exceeds the mental efficiency of direct visual observation.

According to research it takes 20-40 hours of deliberate practice to acquire the skill required to obtain and maintain the “continuous real-time awareness” that allows the OOW to reduce work-load by filtering the amount of information that requires the more precise but higher cognitive workload tools like ARPAs.

Short Summary by Grok
  • Direct visual observation is the primary creator and sustainer of situational awareness
    It delivers continuous, high-bandwidth, real-world data that the brain turns into an accurate, dynamic mental model (Endsley Level 1 → Level 2 → Level 3). Nothing else on the bridge can do this as fast or as reliably in good visibility.
  • The visual picture acts as an intelligent filter
    Once the officer has a solid external mental model, only a tiny fraction of the information on the screens actually needs detailed attention. The eyes tell you what is normal and what is changing; the screens are then consulted briefly and purposefully only to confirm, refine, or investigate the few things that don’t match the visual picture.
    Result: dramatically lower workload, far fewer alarms to chase, and almost no cognitive tunneling.

Relation to Klein’s RPD:

On a ship bridge the fastest, most reliable way to build that recognizable mental model is continuous direct visual observation in real traffic. That’s why junior officers go from “overwhelmed” to “I’ve got this” in four days once they start looking out the window properly — they’re suddenly accumulating the pattern library that RPD depends on.

Each visual encounter adds another prototype to the library. After a few hundred real ones, the brain does the recognition automatically in a fraction of a second — which is why experienced officers often take avoiding action before ARPA even finishes calculating a risky CPA.Screens alone can’t build that library nearly as fast or as richly because they give you processed, symbolic, narrow-field data. You end up doing slow, effortful, analytical decision-making (Endsley Level 2/3 built laboriously from Level 1 fragments) instead of fast, intuitive RPD.

4 Likes
5

How the Brain Builds a World: Visual Perception, Feedback, and the Power of Experience Recent May 2025

From the moment we open our eyes, our brains launch into a lifelong process of building an internal model of the world. It’s an astonishingly rapid and seamless operation: we look, we recognize, we understand.

to see is not just to look, but to know. Our brains don’t simply record the world—they interpret it, reshape it, and build it anew with each passing moment.

Vision is a conversation between memory and sensation, between past experience and present input. And it’s a conversation powered by feedback—a countercurrent stream of information flowing down the visual hierarchy, enriching and refining what we perceive.

The brain rewires itself in real time in response to what’s being seen.

Training mariners to stare at screens creates a huge self-imposed limitation.

6

The Ecological Approach to Navigation: A Gibonian Perspective. - Harry Heft

investigators of navigational processes often uncritically adopt the standard distinction employed by psychologists between perception and cognition. Accordingly, it seems natural to suppose that perception is based on a comparatively simple set of sensory processes,

This paper by Harry Heft was written in 1996, the “standard distinction” between perception and cognition was being overturned. Gibson’s Direct Theory of Perception and Klein’s RPD are based on the more recent view.

The old model (or paradigm) is still deeply embedded in maritime training and bridge design, with all the screens.

7

This paper is about embodied decision making, nothing about marine navigation. But it mentions football

Baseball players with higher batting averages were shown to perceive the ball as larger (Witt & Proffitt, 2005), while American football kickers with higher success rates were shown to perceive goalposts as wider (Witt & Dorsch, 2009).

8

When I took the ship handling course the instructor told the class that if students in the simulator starting using 'body English to assist in controlling the ship it was a sign of being under a bit of stress.

I thought what what kind of knucklehead would try steering the ship by leaning?

A short time later it was my turn in the simulator and I found myself getting set down too close to moored ship. As we finally started slowly eased away I noticed my shoulders were feeling sore. I realized been trying to get the bow to swing clear by using my arms on the handrail to twist the ship.

Asked Grok about embodiment and ship pilots

Yes, a ship pilot leaning into a turn, shifting their weight, or instinctively tilting their body as if that could somehow influence a 100,000-ton vessel in a narrow channel is a near-perfect everyday example of what philosophers and cognitive scientists call embodied knowledge (or sometimes bodied cognition).Here’s why it fits so well:

The rest of the answer here

Yes, a ship pilot leaning into a turn, shifting their weight, or instinctively tilting their body as if that could somehow influence a 100,000-ton vessel in a narrow channel is a near-perfect everyday example of what philosophers and cognitive scientists call embodied knowledge (or sometimes bodied cognition).Here’s why it fits so well:

  1. The knowledge is in the body, not just the head
    The pilot isn’t consciously calculating “if I lean 7° to starboard, the ship will respond 0.3° faster.” The impulse to “body English” the ship comes from years of kinesthetic, tactile, and proprioceptive experience: feeling the helm’s tremor, the slight list of the deck underfoot, the lag between rudder input and the swing of the bow. That felt sense of the ship’s inertia and response has been incorporated into the pilot’s body schema. When stress spikes in a tight spot, the body tries to act directly on the world the way it has learned to act on smaller, more responsive objects (a car, a bicycle, a dinghy).
  2. It reveals the coupling of body and environment
    Embodied cognition argues that cognition isn’t only skull-bound; it’s distributed across body and environment. The pilot and the ship under command temporarily form a single control system. The pilot’s body has internalized the ship’s dynamics so deeply that, in moments of high stakes, the boundary blurs: “my” body now seems to include 300 meters of steel. Leaning or twisting is an attempt to extend intentional motor control into that enlarged “body.”
  3. It’s pre-reflective and non-propositional
    The pilot usually can’t fully articulate why they leaned—they might laugh and call it superstition or habit—but the action emerges from a form of knowing that isn’t verbal or rule-based. It’s the same kind of bodily know-how you use when you shift your weight to balance on a surfboard or “steer” a sled with your shoulders as a kid.

Classic references in the literature:

  • Merleau-Ponty (Phenomenology of Perception) would call this the pre-reflective intentionality of the lived body.
  • Varela, Thompson, and Rosch (The Embodied Mind) use very similar skilled-action examples.
  • Contemporary researchers like Barbara Montero (on expertise) or Dobromir Dotov’s work on human-tool coupling (where people rhythmically move a heavy object together and their brain waves synchronize) cite almost identical cases.

So when you see a seasoned harbor pilot unconsciously throwing “body English” at a supertanker in the Houston Ship Channel or the narrows of the Panama Canal, you’re watching embodied knowledge in its purest, slightly comical, and very human form: the body trying to move the world the way it has learned it can move—only now the world is a floating city, and the body hasn’t quite updated its boundaries.

4 Likes
9

This paper" The Role of the ECDIS on the Development of Situational Awareness – a Study on Grounding Accidents

The paper mentions the workload in high traffic areas but uses “visual” to mean on-screen only, no mention at all of out the window visual.

Here’s an example: "Auditory information has the same value for achieving SA as the visual but often visual information can be revisited on the visual display."

That’s a major oversight.

This is not about using old-timey seamanship but is about understanding fundamental principles with regards to how SA is achieved and maintained. Out the window visual information has it’s limits of course and the various tools available in wheelhouse are valuable and necessary. All available means appropriate etc etc.

10

Anyone who has taught sailing or flying could have told them that. I had students who would happily sail in circles looking 20 feet in front of the boat and we have had new pilots banging off the runway because they were looking right in front of the plane for a century or so.

Standard advice for any n00b on the helm - Do NOT stare at the compass. Get on course and look at the horizon!

11

That’s a good example how visual and instrument information complement each other.

Once the vessel is on course there may be no need to keep eyes glued to the compass. As long as a fixed part of the ship such as a foremast stays aligned with the background, the vessel is not drifting off course.

That visual cue has obvious limits which can be corrected by a glance at the compass.

What about a ship passing between two buoys a mile apart? It’s a more subtle problem, both buoys have apparent motion.

Say there’s a parallel index (PI) line set up on the radar and a track-line on a ENC. Does this task require continuous monitoring on a screen or can it be done visually with an occasional glance at screens used as a cross-check?

1 Like
12

In decent weather I would think you could be 90/10, 90% looking outside and 10% looking in to confirm you are on course.

1 Like
13

The harder problem is one buoy ahead of you and one behind with a crosswind and/or crosscurrent. You need to look behind you to stay in the channel as much as ahead. When that is not physically possible from the helm it gets even harder.

14

Here are two articles that illustrate the two models of how human vision works. The traditional view, what you see is what you get, with no processing (naive realism). This article is from the Nautical Institute, The ECDIS Mindset. , it’s by Andy Norris, the guy that literally wrote the book.

It’s a good article as far as a narrow view using the ECDIS but not for watch-standing in general. The term "visual turns up only as a reference to LOPs. Out the window is referred to as “The view”

image
image468Ă—248 31.7 KB

Here’s the second paper From MarinePublic.com: Mastering Lookout Techniques for Safer Navigation at Sea.

The MarinePublic article is about lookout in the full sense, ECDIS, Radar, ARPA and visual.

Here’s the key point:

Visual observation provides continuous real-time awareness of target movements. Even when not focusing directly on a particular target, peripheral vision detects significant changes. This continuous awareness represents a major advantage of visual lookout that instruments cannot replicate.

The article goes into more detail, has all the expected caveats about reduced visibility, using all available means appropriate and so forth. It’s a fairly long article,

Detection methods should employ all available resources. Visual observation remains the primary technique during good visibility. The human eye detects color, lighting patterns, shapes, and movements that instruments cannot capture.

Over-reliance on instruments creates its own dangers. Inexperienced seafarers often default to RADAR because it provides easy numerical data—CPA, TCPA, course, and speed appear as simple numbers. Yet this convenience comes with reduced information compared to visual observation.

Combining methods provides the most complete awareness. Visual observation should be primary during good visibility, with RADAR confirming and supplementing visual data. During restricted visibility, RADAR becomes primary, but visual checks should occur whenever possible as visibility fluctuates. AIS adds identification and communication capabilities that neither visual nor RADAR observation provides.

1 Like
15

Paper from Endsley Situation Awareness Misconceptions and Misunderstandings (2015)

Endsley gives a good answer to Fallacy #7 but the criticism is still valid.

Fallacy 7: The SA Model Only Represents a Cartesian “In-the-Head” View of the World and Does Not Encompass the Wider Sociotechnical Environment Dekker, Hummerdal, and Smith (2010) states that “situation awareness research typically follows the Cartesian tradition that sets the mind against the world and that maintains that it is meaningful to examine a mind independently of that world (e.g. Endsley 1995)”

Endsely’s model is more “screen friendly” than are Andy Clark’s or Gary Klein’s

I’ve only read bits and pieces but my understanding is that some non-Cartesian, embodied theme writers are Merleau-Ponty, Michael Polanyi and J.J. Gibson

16

What the Nautical Institute calls “The view from the window” J. J. Gibson called an “array”, AKA the real world.

describing an array that exists over a spatial layout and over time, where things move and change, and yet in which persistence of invariant aspects like ground, sky, horizon, and many properties of things, like substance, is specified…Beginning with the world as perceived visually, he called this new physics “ecological optics". Reasons for Realism Selected Essays of James J. Gibson

The optical flow of the ambient array is how people can intuitively and implicitly judge their position as they move, for example the bridge team in the wheelhouse while in pilotage waters.

Learning to use optical flow is called perceptional learning, which requires watching over time, not just quick glances out the window.

17

This paper from May, 2025.

Key takeaways

  • Neurons in the ventral visual cortical pathway help our brains make visual sense of the world.
  • A new study finds that these neurons adapt moment to moment to the input they receive.

The view outside is not just another tool like ECDIS / ARPA that can be read at a glance.