771
1 INTRODUCTION
In Japan, to obtain a driver’s license, most people take
over 24 hours of classroom instruction and over 31
hours of practical instruction, as well as examinations.
To obtain a first-class boat operator permit, you must
have over 24 h of classroom instruction as well as 12 h
of practical instruction. In addition, a renewal class is
necessary every five years for both driver’s license and
boat operator permits. In Japan, people must pass an
exam for a small boat and obtain permission from the
boat’s operator. People study written lessons for more
than 24 hours with on-board training, and they
maneuver small ships with less than 20 gross tonnages.
Operators must take lessons for the renewal permits
every five years. It is difficult for operators to practice
with experts after passing an examination. The novices
cannot steer well, assess weather information, do
safety-checks, and look out for safety navigation in case
of “something wrong.” I analyzed the movement, time,
and frequency of gaze-plotting during steering due to
being a novice, similar to experts. None of the groups
used glasses. In this study, the gaze plot was defined
as the point at which the operator looked hard. I
considered the operators’ improved performance,
what they were interested in, and where they looked
from gaze plotting. In the European Union, there is no
uniformity in obtaining a permit to operate small
vessels, except for fishing boats; in many cases,
especially small boats and yachts, a permit for a small
boat is not required. [1] In this study, the operators
were placed on eye-tracking devices. My purpose is to
determine where and when operators steer and look
out during the maneuvers. Figure 1 shows a small
experimental boat,” AKEBONO,” belonging to NIT,
TOBA College in JAPAN. (LOA= 6.62 m, B= 2.56 m,
G/T= 2.8 tons)
Analysis of Boat’s Operator Action Using Gaze-data
N. Yoshida
1
, Y. Imai
2
, K. Minami
3
& H. Seta
2
1
National Institute of Technology, Toba College, Toba, Japan
2
Tokai University, Tokyo, Japan
3
Tokyo University of Technology, Tokyo, Japan
ABSTRACT: In Japan, obtaining a small boat operator’s license requires both theoretical and practical training, as
well as mandatory renewal courses every five years. However, many operators have limited practical experience,
making safe navigation a challenge for beginners. This study aims to analyze the differences in gaze patterns
between novice and experienced operators using eye-tracking technology, highlighting key behavioral
distinctions. Eye tracking data was collected from 14 operators, including novices and experienced individuals.
Gaze movements were recorded in a 360-degree field of view to measure gaze duration and frequency. Heatmaps
and grid-based Areas of Interests (AOI) were generated to analyze patterns of gaze shifts and movement
directions. The results indicated significant differences in gaze behavior, field of view, memory retention, and
decision-making between novice and experienced operators. Experienced operators process information more
efficiently, leading to safer navigation, whereas novices exhibit narrower focus and slower reaction times.
Integrating eye tracking data into training programs can enhance beginners’ navigation skills.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 19
Number 3
September 2025
DOI: 10.12716/1001.19.03.09
772
1.1 Defined Lookout
Lookout is defined by the international and domestic
regulations in each country. Look-out by COLREG is to
the COLREG, seafarers aboard every vessel shall
always maintain a proper look-out by sight and
hearing as well as by all available means appropriate
in the prevailing circumstances and conditions to make
a full appraisal of the situation and the risk of collision.
As mentioned above, the role, importance, and job
descriptions of watchkeeping in merchant vessels were
explained. However, the implementation method is
explained only in terms of international conventions,
and no specific examples are provided. [2] Although
not found in general merchant vessels, the following
description of the lookout method in the U.S. Navy was
found. It explains that effective visual searching does
not come naturally; a lookout must learn through
practice and allow your eyes to move in short steps
from object to object, and you can really see what is
there. [3][4]
1.2 Boat on Instruction Manual in JAPAN
We researched how operators should LOOK OUT
instruction manual. The instruction manual for the
small boat renewal course states that the operator must
always maintain a proper lookout, in accordance with
maritime regulations, to determine whether a collision
is imminent or to assess the vessel’s surrounding
conditions. Looking Out is always performed at sea,
such as anchoring, stopping, and sailing. A safety
Check was performed before performing the different
actions. Operators are required to monitor the vessel’s
current position while anchoring, stopping the engine,
and sailing. Their actions include starting the engine,
adjusting speed, changing course, and other
maneuvering tasks. Operators learn lookout
procedures from instructional materials designed for
instructors, which emphasize that "the captain must
check the surroundings and maintain a lookout in
accordance with maritime regulations, considering the
risk of collision with other vessels." Maintaining a
proper lookout is essential during mooring, anchoring,
and sailing. A safety check must be conducted before
any maneuver and includes confirming the vessel’s
current position while at anchor, with the engine
stopped, or under way. The movements of the
operators are defined as” Ahead Engine”, ” Alternative
Speed,” and” Changing Course.”
Figure 1. Experimental small boat “AKEBONO”
1.3 License and Training in JAPAN
In Japan, to obtain a driver’s license, most people take
over 24 hours of classroom instruction and over 31
hours of practical instruction, as well as examinations.
To obtain a first-class boat operator permit, you must
have over 24 h of classroom instruction as well as 12 h
of practical instruction. In addition, a renewal class is
necessary every five years for both driver’s license and
boat operator permits. In the classroom, trainees study
the responsibilities of a captain, traffic rules, and
general maneuvering; in practical instruction, they
receive training about checks before departure, basic
handling, and advanced handling.
2 METHOD OF ANALYSIS
Eye tracking was used as a method to measure the
behavior of the operators. Eye tracking was chosen as
a means to investigate the direction and object that the
operators were paying attention to. Previous studies of
automobiles and aircraft using eye tracking have
investigated the field of view, experience, and eye
tracking movement of automobile drivers.[5][6][7] As
eye trackers become small and remote, we could easily
measure on small vessels, and Sematic Gaze Mapping
allows the operator's behavior to be visualized. [10]
The measurement in this study is the method of the
following.
1. Twenty operators participated in the
measurements. There were two groups: novices and
experts.
2. Operators wearing gaze-plotting tools maneuver at
the port.
3. The gaze plot of the operators draws a view of 360°.
4. Time and number of gaze plots were calculated
from the recorded data.
5. The operators heard about their movements from
the recorded data.
6. The heat maps and grid AOI were calculated from
the gaze-plotting data.
7. The order of the gaze plotting was calculated.
The directions from the small boat were” front,””
left, ”” left back,”” right,”” right back,” and” back,”
every 60 °, as shown in Figure 2. The novices had the
First or Second Permit of the Boat’s operator within
two years, and the experts had the First Permit of the
Boat’s operator with their duties. The training teacher
stood for safe navigation as she always steered during
the measurement.
Figure 2. Boat’s direction in this experiment
3 RESULT OF ROUTE AND TIMELAPSE
Position data were obtained from GPS data. Time-lapse
data were calculated from gaze-plotting data.
773
3.1 Introduction to the entire route
The route in this measurement was shown in Figure 3
and was explained below.
1. The operators check the crossing of the municipal
ferry and the same- and opposite-course vessels
after starting. There is an island on the port side.
2. The operators check the municipal ferry from port
side to port side in narrow areas halfway around
this route.
3. The area to end around this route is a larger area in
which the operators check the position of the other
ships.
4. Figure 4 and Figure5 showed the result that one of
them steered when and for what purpose.
Figure 3. Testing route in this experiment
Figure 4. Heat map in Novice and Expert
Figure 5. Total of gazing time and number of gaze plotting
3.2 Introduction to area SA: from point S to point A
The SA area lies between points A and point B in
figure 3. This area has the following characteristics.
1. The SA is a route for ferries and other ships. There
are many crossings, the same course, and opposite
course ships.
2. The experts navigated the center of the sea area due
to the island on the port side.
3. Novices navigate the starboard side of the sea area.
The experts observed other ships rather than
shallow areas, and the novices observed shallow areas
rather than other ships.
3.3 Introduce to the area AB: from A to B
Area AB lies between points A and B in figure 3. This
area has the following characteristics.
1. Area AB is a narrow, curved route. The operators
navigated near SAKATE Island.
2. The experts alternated faster than the novices off the
Maruyama saki lighthouse and the novice alter
quickly. The experts had enough time to steer in the
narrow sea and navigate the port side off SAKATE
Island.
3.4 Introduce to the area BG: from B to G
1. The operators move straight between B and G and
search for Point G.
2. Operators navigate the best route. Each operator
chooses a different route.
Operators steered to the surroundings of the large
sea. The heat maps of the experts and novices were
compared. I cannot understand the characteristics of
the two groups because of the large red area, where the
operators watch the gaze point for a long time. I made
the Grid AOI the operator’s gaze point area.
− The novices had a smaller gazing range in the upper
direction. The experts gazed at three areas from the
front, while the novices gazed at two areas from the
front.
− The novices gazed at the nautical instruments for a
longer time than the experts. Boat operators require
physical standards .
− Even if the visual acuity of one eye is less than 0.5,
the other eye must have a visual field of 150 degrees
or better in both directions and a visual acuity of 0.5
or better.
− Visual acuity must be 0.5 or better in both eyes.
As the inspection standards were seen, the
standards for boat operators are based on eyesight.
Therefore, it is important for the operators to gaze over
a certain range and observe what they can see.
According to the permit standard, for an operator to
progress, or she must be able to see over a large area
and have eyesight to obtain the necessary information
for navigation. Experts are more skilled than novices
because they see a larger area.
4 RESULT OF GAZING POINTS.
The heat map Results are shown in Figure 4. The color
was purple when the gazing time was long, and blue
when the gazing time was short. The novices and
experts are shown in purple on the front side. This
shows that they looked forward to the front most often.
Since the heat map was not provided numerical results,
the area was divided the area of the eyesight range into
a grid to evaluate the gaze time and frequency in each
direction. The number of gazes and the gazing time
were calculated for each area. The characteristics of
each operator are described in the following
subsections.
774
4.1 Result of gazing time and number of gazing points.
Since there were no significant differences between the
two groups in gaze duration and the number of gaze
points in each direction, except for the “Port side,” the
number of gaze transitions between directional pairs
were analyzed, as shown in Figure 5 and Table 1. Table
1 confirms the absence of significant differences
between the practitioner (faculty) and novice groups.
During maneuvering, operators primarily directed
their gaze toward the front, starboard, and other
surrounding areas; however, in an actual operation, the
operator is not looking at each side only, but rather, in
relation to each other, monitoring how the
surrounding situation changes over time. Because the
situation changes from moment to moment, the subject
must pay attention not only to the front, but also to
each direction. Therefore, in this analysis, in addition
to the calculation of the number and duration of the
gazing point, the number of times, and for each
direction, the gazing point that moved between the
seven directions was calculated.
1. Novices and experts performed the same number of
movements from A to B.
2. The gaze point moves from the front side to the
other side and from the other side to the front side
of the AOI.
3. There were many instances where the expert’s gaze
point moved from the port side to the front side.
4. Numerous times, the novice’s gaze point moved
from the nautical instrumental side to the front.
5. The experts looked forward to the nautical
instrumental side for a shorter time than the
novices. The gaze point of the expert moved
between the port and the front sides many times.
6. One of the expert groups with a license for practical
instruction on boats maneuvered once a week on
another vessel using this measure.
7. To understand the operator's determination, the
results of the time-lapse gaze data were used to
measure the operator's intention when gazing in a
direction other than in front of the operator.
5 RESULT OF HEARING EVERY MANEUVER
The expert kept watching the wide area and explained
the object that he looked at. They were able to
understand their behavior while maneuvering and
explain their performance. They remembered their
maneuvers, other ships, surroundings, weather, sea
conditions, the objects on the sea, and their feelings in
Figures 6, 7, and 8. Novices carry out lookouts for small
areas, and they cannot look at other ships, buoys, or
objects on board. They cannot decide whether other
ships, buoys, or objects are dangerous to their boats.
They believe that boat operators have many tasks, so
novices do not notice something that is too dangerous
or prevent it. Many novices cannot understand the
behavior they perform or explain their maneuvers.
Table 1. Significant difference each direction
Figure 6. Behaver of the novice’s horizontal gazing point and
his purpose
Figure 7. Behaver of the expert’s horizontal gazing point and
his purpose
775
Figure 8. The purpose of operator lookout to port side and
starboard side
6 DISCUSSIONS
The table shows the difficulty levels for each area: SA,
AB, and BG. The characteristics of each area are as
follows.
1. A larger value is more difficult to obtain, whereas
smaller values are easier to obtain.
2. Evaluation items included information on the
navigable area, weather, sea conditions, and own
vessel speed, and on other vessels: number,
congestion, size, type, and distance from the vessel.
3. Because there are linear vessels and ports in the SA
area, inland vessels often arrive and depart from
ports. The largest vessels were in the BG area
because the BG area has the largest SG area. The
characteristics of the novices and experts in
maneuvering during hearing research are
presented in Table 2.
From 7 persons to 10 persons applied d to “〇”,
from 4 persons to 6 persons applied to “△”, and from
0 persons to 3 persons applied to “×” in every content.
The contents were as follows: remembering how I
maneuvered, explaining why I performed that
maneuver, remembering where I altered the course
after this voyage, prioritizing maintaining speed,
checking something unrelated to the maneuver, asking
about that course during the maneuver, and checking
the surroundings during the maneuver. The experts
were able to remember and explain the maneuvers and
reasons for performing them. First, they responded
that they could hardly remember the event; however,
when they viewed the video of the maneuver, they
recalled their behavior and were able to provide an
account of the event and their intentions during the
measurement. In the beginning, they responded that
they could hardly remember the maneuver, but when
they watched the record of the maneuver, they
explained their behavior and were given an account of
the event and their intentions. However, after that
maneuver, the novices could not remember much
about the maneuver or explain their steering.
According to them, they could avoid objects with the
steering wheel but could not remember the maneuvers.
The experts can provide information about their ships
and surroundings. However, the novices had little
experience. Novices do not have a good order of
priority for steering information to understand their
navigational surroundings. Although they could
assess the state of the boat at the time of measurement,
the necessary information at the time was not utilized.
7 CONCLUSION
This study analyzes operators’ maneuvering and
behavior using a gaze-tracking device. Gaze data were
collected from 14 operators licensed for less than two
years and eight faculty members licensed for more than
three years while they navigated a designated route.
Because there was a significant difference in gaze
duration and frequency when looking to the left, the
front direction was divided into finer grids than those
used in the heat map. Time and frequency of gaze were
then analyzed within these detailed areas. A
comparison of gaze point distribution showed that
faculty members had a wider range of gaze points
lasting more than one second. In the interview with the
operators on viewing the measurement data, the
novices did not remember their behavior and
maneuvering. When reviewing the recorded footage,
the experienced operators were able to explain their
actions and intentions, whereas the novices were not.
In the analysis of the sequencing gazing point, the
experts gazed at the port side more frequently than the
novices, and there were significant differences in the
time and number of gazing points. In the analysis of
the gazing point, it was found that the experts gazed at
the instruments more frequently, while the novices
gazed at the instruments for a longer time. This is in
accordance with previous research on high-speed sea
navigators and car drivers.[9][10] Students required
more time to understand the information. When the
measurement data were reviewed for each operator, it
was found that some novices had prior experience
maneuvering boats in their personal lives. Compared
to novices who had not operated a boat since obtaining
their license, these individuals focused more on the
front, instruments, starboard, and port sides. They also
spent more time looking at, and looked more
frequently toward, the starboard side than the port
side. The experienced operators, although unfamiliar
with the specific boats used in the study, had navigated
the measurement area with other vessels. As a result,
they exhibited a narrower range of gaze and looked in
fewer directions, likely because they already knew
where to focus their attention while navigating.
Among both novices and experts who had no prior
experience with the measurement area or the vessels
used, there was no clear pattern distinguishing those
who gazed for extended periods from those who made
multiple short glances. This variation appeared to
depend on individual tendencies rather than
experience alone. Novice and expert groups were also
identified. The novices and experts were compared.
There were significant differences in gaze time and
frequency on the starboard side and the starboard back
side between novices and experts. The area where the
novices looked out for less than a second was
calculated. The direction in which operators looked out
from anywhere to another was analyzed because there
was no definite difference in significant gaze at time
and frequency. The analyzed gaze sequence was used
to check the direction in which the operator looked at
the sequence. The method counts when the gaze moves
from one of seven directions (port back side, port side,
front side, starboard side, starboard back side, back
776
side, and nautical instrument side) in another
direction. For example, if the gazing point moved from
the front to the port side, it was counted once. If the
gazing point moved to the front, it was not counted. A
comparison of the median and average values from the
gaze sequence analysis for novices and faculty
members revealed that the most frequent gaze
transitions in both groups were from the front to the
port side and from the port side back to the front. These
were followed by transitions between the front and
starboard sides. While the number of times both
groups looked from the starboard side to the front side
and front side to the starboard side, both the average
and median values for the front side to the port side
and port side to the front side were more than 10 times
higher for the experts, indicating that the experts were
more likely to pass their gaze points in these directions.
However, because the operator’s intention for the
maneuver could not be determined from these
measurement results, an interview was conducted to
confirm why this action was performed by viewing the
recorded video from the operator’s gaze point at the
time of measurement with the individual. The
interview research found that most experts could
remember their maneuvering actions and explain the
intent of their actions when they watched the
recording, whereas most novices forgotten the intent of
their actions and could not explain them. The other
findings were as follows:
1. The experts crossed a safe distance, did not
suddenly change courses, and maintained a clear
course in the other vessel.
2. Novices have a long time to consider and change
courses by avoiding maneuvering after
approaching them.
3. Neither group was forced to stop during
navigation.
4. If the boat collided with another boat would make
a collision, the expert performed a change (slowing
down or changing course) of the vessel in an early
situation.
5. Novices are slower and take longer to make
decisions than experts because they try to handle
relationships with multiple boats and objects
simultaneously.
Groups of novices and experts were distinguished.
The novices and experts were compared. Novices and
experts gazed at the time and frequency on the
starboard side and the starboard back side. The area
where the novices looked out for less than a second was
calculated. The operators looking out from anywhere
to another were analyzed because of the lack of definite
significance of gazed time and frequency. The analyzed
gaze sequence was used to check the direction in which
the operator looked at the sequence. The method
counts when the gaze moves from one of the seven
directions (port back side, port side, front side,
starboard side, starboard back side, back side, and
nautical instrument side) to the other seven directions.
For example, if the gazing point moved from the front
to the port side, it was counted once. If the gazing point
moved to the front, it was not counted. Comparing the
medians and averages of the gaze sequence analysis for
the novices and faculty members, the most frequent
gaze points passed for both groups were from the front
side to the port side and from the port side to the front
side, followed by the front side to the starboard side
and the starboard side to the front side. While the
number of times both groups looked from the
starboard side to the front side and front side to the
starboard side, both the average and median values for
the front side to the port side and port side to the front
side were more than 10 times higher for the experts,
indicating that the experts were more likely to pass
their gaze points in these directions. However, because
the operator’s intention for the maneuver could not be
determined from these measurement results, an
interview was conducted to confirm why this action
was performed by viewing the recorded video from the
operator’s gaze point at the time of measurement with
the individual. The interview research found that most
experts could remember their own maneuvering
actions and explain the intent of their actions when
they watched the recorded views, whereas most
novices forgotten the intent of their actions and could
not explain them. The other findings were as follows:
1. The experts crossed a safe distance, did not
suddenly change course, and maintained a clear
course to the other vessel.
2. The novices had a long time to consider and change
course by avoidance maneuvering after
approaching.
3. Neither group was forced to stop during
navigation.
4. If one boat and another boat collided, the expert
performed a change (slowing down or changing
course) of the vessel in an early situation.
5. Novices are slower and take longer to make
decisions than experts because they try to handle
relationships with multiple boats and objects
simultaneously .
The gazing time, duration, and sequence tended to
be the same for novices and experts. The novices all
had permits, so they passed the exam as boat operators.
However, if the experts performed movements from 90
to 100 points, the novices performed movements from
60 to 100 points.
ACKNOWLEDGEMENT
In this study, boat operators looking to operate their boats are
analyzed and discussed. The location of the gaze point,
number of gaze points, total gaze time, and gaze sequence
were calculated. There were no differences in the gaze-point
analysis between students and faculty members, and
interviews were conducted, which revealed differences. I
thank the Japan Coast Guard, our boatyard staff, the Toba
Maru crew, faculty members, our students, and external
collaborators who consulted on the area during the boat
maneuvering measurements, the Bioethics Committee for
their permission to conduct the research, and my supervision
of students for their cooperation regarding the content of this
study.
REFERENCES
[1] J. L. Alonso-Enriquez, F. Piniella, J. I. Alcaide, “The Need
for Unification and Harmonization in a Maritime
Licensing Standard for Recreational Craft”, TRANSNAV,
2020
[2] House. D. J, “SEAMANSHIP TECHNIQUES 3rd edition”,
2012 Andrew. J. May, Tracy Ross, Steven. H. Bayer, 2012,
Drivers’ Information Requirements when Navigating in
an Urban Environment, The Journal of Navigation, The
Royal Institute of Navigation
777
[3] CENTER FOR SURFACE COMBAT SYSTEMS. U. S.
Navy, “Lookout Training Handbook”, NAVAL
PERSONNEL DEVELOPMENT CENTER, Virginia,
United States of America, 2007
[4] Department of Homeland Security United States Coast
Guard, “Navigation Rules and Regulations Handbook”,
Washington, D. C , United States of America, 2020
[5] T. MIURA, “Visual search in intersections-An underlying
mechanism”, IATSS Research, vol. 16, pp.42-49, 1992
[6] D. Crundall, G. Underwood and P. Chapman, “Driving
Experience and the Functional Field of View”, Perception
28 (9), pp.1075-1087, 1999
[7] G. Underwood, P. Chapman, J. Underwood and D.
Crundall, “Visual attention while driving: sequences of
eye fixations made by experienced and novice drivers”,
Ergonomics, vol. 46, pp.629-646, 2003
[8] D. Martinez-Marquez, S. Pingali, K. Panuwatwanich, R.
A. Stewart and S. Mohamed, “Application of Eye
Tracking Technology in Aviation, Maritime, and
Construction Industries: A Systematic Review”, sensors,
MDPI, vol. 21, no.13, 2021
[9] F. Forsman, A. Sjörs-Dahlman, J. Dahlman, T. Falkmer
and C. Lee, “Eye tracking during high speed navigation
at sea: Field trial in search of navigational gaze
behaviour”, Journal of Transportation Technologies, vol.
2, pp. 277-283, 12
[10] B. Muczyński, M. Gucma, M. Bilewski and P. Zalewski
Using, “Using Eye tracking data for evaluation and
improvement of training process on ship’s navigational
bridge simulator”, Scientific Journals, Maritime
University of Szczecin, vol. 33, no.105, pp75-78, 2013
