889

1 INTRODUCTION

The works of International Maritime Organization

(IMO) and the European Commission to develop and

implement regulations governing the safe navigation

of autonomous vessels are still in progress. At the

105th session of the IMO International Maritime

Organisation Maritime Safety Committee (MSC) in

April 2022, the works on developing assumptions

governing the implementation of the work program

on MASS (Maritime Autonomous Surface Ships)

began and it was expected that the MASS Code would

enter into force on January 1, 2028.

On the 108th session [20], in May 2024, MSC

agreed to the revised Road Map for the MASS code

development, including introduction of amendments

to SOLAS Convention (International Convention for

Safety of Life at Sea), adoption of a mandatory MASS

Code in 2028 and its entry into force on January 1,

2032. This code will regulate and solve problems of

technological and legal nature, transport safety,

including operational reliability of MASS Complex

Technical Systems, navigation, ship operation,

monitoring, Search and Rescue (SAR), legal aspects,

cooperation between MASS and the Remote

Operations Centre (ROC), and ROC operator

competencies.

One of the complex issues difficult to formalize is

the relationship between human responsibility and

automation of processes related to safety of ship,

cargo and marine environment, when MASS is

monitored and controlled from the shore by ROC

operator. The competencies of the operator are

covering the following areas:

− operation with legal liability for the ship and

cargo,

− voyage planning, remote navigation

− new collision avoidance regulations,

Situational Awareness in Autonomous Shipping – Ship

Domain in Remote MASS Operation

T. Abramowicz-Gerigk & Z. Burciu

Gdynia Maritime University, Gdynia, Poland

ABSTRACT: The introduction of Maritime Autonomous Surface Ships (MASS) in maritime transport creates

new challenges that did not previously exist in the case of manned ships, and changes the approach to voyage

planning, implementation and monitoring. MASS is not only supposed to be more economical, but also

contributes to transport safety and environmental protection, while limiting the impact of the human factor.

Taking into account the assumptions of the International Maritime Organization, the implementation of a MASS

voyage, supervised by the operator of the Remote Operations Centre (ROC) will require a high level of

situational awareness. The paper discusses the determination of the MASS safe navigation domain by ROC

operator making decisions under risk conditions. It is expected, that according to Kahneman and Tversky's

prospect theory, the enlargement of MASS domains may result in an increase in human-induced navigation

hazards, especially in restricted areas.

http://www.transnav.eu

the International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 18

Number 4

December 2024

DOI: 10.12716/1001.18.04.15

890

− MASS safety and security monitoring,

− preventing and responding to cyber hazards,

− cooperation with SAR services in rescue

operations.

In the absence of international rules and standards,

the alternative way of ensuring safety was introduced

by EMSA (European Maritime Safety Agency) to

create the common safety levels and enhance

harmonization in the analysis of preliminary MASS

designs. Det Norske Veritas (DNV) was contracted by

EMSA to develop a Risk Based Assessment Tool

(RBAT) [15].

In the MASS hierarchical control structure

presented by Thombre et al. [16], data related to

situational awareness, available for Shore Control

Centre operator, comes from Integrated Automation

System, which uses data from MASS Autonomous

Navigation System. In the opinion of experienced

Ship Masters, comparing the manned and remote ship

operation, situational awareness (SA) based on the

data from Integrated Automation System causes the

operator to change the boundaries of safety measures.

The names introduced by various authors for the

land-based MASS operations centre such as Remote

Control Centre introduced by DNV for remotely

controlled ships or Shore Control Centre proposed by

Thombre et al. [16] fall within the general concept of a

Remote Operations Centre ROC.

The paper discusses one of the most important risk

options related to situational awareness of ROC

operator, which is the ship safety domain.

2 SITUATIONAL AWARENESS IN

AUTONOMOUS SHIPPING

Human situational awareness in autonomous

shipping is the ability of the Remote Control Centre

operator to perceive, understand and predict changes

in the actual situation in order to conduct a safe

voyage, based on signals received from the Integrated

Automation System.

The key elements of MASS and Remote Control

Centre are presented in Figure 1.

Figure 1. Key elements of MASS and Remote Operations

Centre

The automatic situational awareness system,

devoted to the monitoring and interpretation of own

ship surroundings, manages and utilizes the

information from onboard systems e.g.: AIS, ECDIS,

GNSS, radar, lidar, IR, cameras, speed log, echo

sound, gyro compass, microphone, thermometer,

anemometer, and inertial measurement unit. Based on

AI (Artificial Intelligence) algorithms, the combined

sensor data - multi-sensor perception system -

provides the required situational awareness [16].

In remote operation, according to IMO, MASS is

managed by Remote Operations Centre, depending

on the level of autonomy. The situational awareness

in this case is based on operational procedures, ROC

operator training, his personal attitudes and skills. SA

in the MASS remotely managed voyage should help

avoid accidental events resulting in disruptions.

The key objective of the situational awareness is to

prevent errors that may arise at one of three levels:

perception, understanding and prediction [14]. For the

operator in the Remote Control Centre, as opposed to

the Ship Master on board the ship, an additional area

of situational awareness is created due to automation,

remote sensing and remote management.

The implementation of a safe MASS voyage will

involve a high level of situational awareness, which

will enable the operator to counteract possible

hazards and predict their status in the technical and

commercial operation of the ship.

The three levels of situational awareness

perception, understanding and anticipation are

presented in figure 2.

Figure 2. Situational awareness of operator in MASS

operation

Level 1 of situational awareness - perception –

means receiving and recognizing important

information, enabling sea voyages, including proper

visualization of operational, technical, environmental

and navigational parameters presented by the MASS

digital twin and information from vision sensors.

Level 2 - understanding – is the level of

understanding and interpreting information obtained

at level 1.

891

Level 3 - anticipation – incudes forecasting the

future status of sea voyage by ROC operator.

The operator, based on the information from

MASS Integrated Automation System and signals

from the ship digital twin, is responsible for making

decisions on restoring the deviations of system

parameters from permissible values to the required

range [11]. The perception of navigational and

technical information provided and the

understanding of its meaning is required to respond

quickly to a situation. The obligation of the operator is

to implement the road map of the sea voyage in case

of II, III, and also IV degree of MASS autonomy in

emergency situations.

Currently, the preparation of a manned ship

voyage plan must meet guidelines, included in IMO

Resolution A.893(21) and dependent on situational

awareness of the crew onboard. The plan may be

changed any time the situational awareness changes

or when it is possible to find a safer option.

The commercial platforms delivering dynamic

voyage planning, with multi-objective optimisation,

related to real time weather observations and ship

performance models, based on artificial intelligence

methods e.g. Wayfinder [22], SeaPerformer [21] or

Short Horizon Planner for collision avoidance,

adaptable as tool for decision support or automated

route deviations, presented by Enevoldsen et al. [5],

have been developed. Several studies and projects of

autonomous ships are at various levels of

development and application including fully

autonomous ships in operation [16,19].

Because the actual level of situational awareness

influences the necessary safety measures taken by

both autonomous navigation algorithms and human

control, these should be tailored to the task and take

into account the operator's personal skills and

confidence in the reliability of the AI. In most cases,

many years of Ship Master’s experience on board is

the reason for less confidence in individual devices of

decision support systems [3].

Table 1 presents the relationship between

situational awareness and confidence, originally

presented by Endsley & Selcon [6].

Table 1. Relationship between situational awareness and

confidence [6]

________________________________________________

Situation Awareness

Good Poor

________________________________________________

High Good outcome Bad outcome

Low Do nothing Satisfactory

Confidence level ineffectual outcome

Delay

________________________________________________

The person having Poor SA and high level of

confidence, presents the worst outcome, making

wrong decisions and giving false confidence to other

personnel. This is the most dangerous situation of all

the situations presented in table 1. The good outcome

is dependent on both good situational awareness and

high confidence level.

This important issue should be carefully

considered to avoid over-reliance, misunderstandings

or conflict between the operator and decision support

systems at the three levels of SA.

Aylward et al. [2] using Advanced Intelligent

Manoeuvring (AIM), developed by Wärtsilä, explored

i.a. how the decision support system can influence the

safety of navigation and the role of the operator in

routine ship traffic situations. When a MASS operator

has poor situational awareness due to limited trust in

information from all decision support systems, the

MASS Integrated Automation System and the digital

twin, the solution to ensure a safe and satisfactory

outcome is a delay, which may result in a decision to

reduce the ship's speed and increase the ship's

domain.

3 MASS DOMAIN DETERMINED BY ROC

OPERATOR – HANNEMAN & TWERSKY

PROSPECT THEORY

The concept of a safe navigational domain around the

ship was created to determine the safety of ship traffic

in restricted areas. The domain can be defined as the

area around the ship, dependent on vessel parameters

and traffic situation, in which no other objects should

be present [9,13].

Various methods are used to determine the ship's

domain based on analytical techniques and

simulation, combined with statistical methods and AI.

The factors affecting domain parameters include i.e.

navigational area characteristics and ship length. A

statistical study of ship domains with a method for

determining the safe passing distance of a ship was

developed by Goodwin [8].

The empirical data show the domains defining the

navigator’s comfort zone with dimensions of 4.5 ship-

lengths in front of the ship and 3.5 ship-lengths

behind [9,12]. This shape can be extended into a

super-ellipse [5] or circle (Figure 3).

Figure 3. Ship domains: A - manned ship domain, B -

remotely controlled MASS domain with the coverage areas

of navigation sensors

The example of ship domains determined by Ship

Master and ROC operator as safety zones, presented

in Figure 3, show the approximate boundaries of ship

safety navigational areas from their perspective

892

The larger domain is related to the operator's

decision making under risk conditions. The

assessment of a risky situation depends more on the

reference point from which we calculate loss or gain,

than on their final values. Losses always seem to be

greater than gains. A major loss causes permanent

risk aversion.

Limited situational awareness and lack of

understanding of the situation, trigger emotions that

have a huge impact on risk assessment. It is fear that

causes the perceived risk to exceed the actual risk,

while euphoria, in turn, reduces the perceived risk

level of a given situation [1,4].

Decisions made under conditions of risk were

analysed by Kahneman & Tversky [10]. In unclear

conditions, choices become irrational and personality

significantly influences decisions. People estimate

probabilities of individual events overestimating

medium and small, and underestimating large

probabilities of failures. The Kahneman & Tversky

prospect theory introduced the decision weights

instead of the probability functions.

ROC operator having Ship Master skills and

experience, managing MASS in restricted navigation

zones, will designate the MASS domain, which,

assuming the prospect theory of Kahneman &

Tversky, will be burdened with risk aversion. This

results in the larger domain “B” than the domain

designated by the Ship Master on board the manned

ship ”A” (Figures 3, 4).

Figure 4. Domains of a ship with length >150 m in Traffic

Separation Scheme: A – computed on the basis of statistical

AIS data [13} and B enlarged domain of the remotely

controlled MASS

The domains observed in traffic situations show

different domains, e.g. for ships with length greater

than 150 m, in Bornholmsgat TSS (Traffic Separation

Scheme) the domain dimensions computed on the

basis of statistical data from AIS, were about one NM

in front and behind ship [13] (Figure 4).

By increasing the area of the MASS domain, ROC

operator will expand the situational comfort zone,

increase the time reserve, eliminating the time

pressure for making rational decisions, and at the

same time will gain time to obtain additional

information. Enlarging the domain will cause the

operator to maintain a safe MASS speed.

4 DISCUSSION AND CONCLUSIONS

It is expected that autonomous shipping will reduce

the impact of the human factor on the risk of failure.

One of the elements ensuring larger comfort zone

related to the situational awareness of Ship Master, is

creating own ship domain, providing the availability

of time to make decisions and avoiding hazards.

Full information will enable the voyage to be

maintained at the required level of safety. The

perception of navigational and technical information

provided and the understanding of its meaning

requires to respond quickly.

Possible misunderstanding of the ship's situation

may result from delays in communication and

decision-making resulting from long lead time to

approval - throwing people out of the loop, as well as

errors resulting from the "transfer effect" when the

operator is involved in simultaneous monitoring of

several ships and assesses their response based on

signals from devices without feeling the ship's

reaction.

Taking into account the research on conventional

ship domains [13] and using Kahneman and Tversky's

prospect theory in relation to the dimensions of the

MASS domain, in order to avoid domain violations

and hazardous situations, the parameters of

navigation areas e.g. traffic separation systems, traffic

lanes and separation zones should be reconsidered.

The important issue is that the instruments used

onboard MASS and in Remote Operations Centre will

meet performance standards [17] and whether the

training and development of skills of seafarers, and

ROC operators will be of high priority for IMO [2, 18].

The problem is more complex when the MASS

operation is carried out in autonomous integrated

transport chain [7], by several operators, at individual

stages of the voyage and when an operator controls

several autonomous ships. This is also related to the

cooperation between VTS and ROC.

Baldauf et al. [3], who presented simulation study

on the VTS collaboration with ROC, concluded that

VTS operators took earlier actions, than in their usual

practice, to establish communication with MASS,

which means less confidence. The biggest concern was

related to the uncertainty of MASS remote operator’s

situational awareness and conclusion was that the

new technologies enhancing SA should be applied.

The same data should be available in ROC as is

available on board the ship and the skills of the ROC

operator should be consistent with the STCW

893

requirements for shipmasters and officers of the

watch. However, in the case of many MASS units in a

traffic separation zone, a situation may arise that the

dimensions of the traffic separation zones may be

insufficient.

If the presented theses of the paper regarding the

enlarged MASS domains by ROC operator are

realized, it may result in an increase in navigation

hazards caused by the human factor, especially in

restricted areas, in traffic separation zones.

ACKNOWLEDGEMENT

This work was supported by the project of Gdynia Maritime

University No. WN/2024/PZ/08.

REFERENCES

[1] Abramowicz-Gerigk T., Hejmlich A. Human factor

modelling in the risk assessment of port maneuvers.

TransNav The International Journal on Marine

Navigation 9,3, pp. 427—433, 2015.

[2] Aylward K., Weber R., Lundh M., MacKinnon S. N.,

Dahlman J. Navigators’ views of a collision avoidance

decision support system for maritime navigation. The

Journal of Navigation 75: 5, 1035–1048, 2022.

(doi.org/10.1017/S0373463322000510).

[3] Baldauf, M., Rostek, D. Identify training requirements

for remote control operators of maritime autonomous

ships. Proceedings of 18th International Technology,

Education and Development Conference 2024.

(doi:10.21125/inted.2024.2036).

[4] Burciu Z. Reliability of rescue action. Warsaw University

of Technology Printing House. Warsaw 2012.

[5] Enevoldsen T.T., Blanke M., Galeazzi R. Sampling-based

collision and grounding avoidance for marine crafts.

Ocean Engineering 261, p. 112078. (doi:

10.1016/j.oceaneng.2022.112078.

[6] Endsley M. R., Selcon S. J. Designing to Aid Decisions

Through Situation Awareness Enhancement. 2nd

Symposium on Situation Awareness in Tactical Aircraft

Patuxent River, MD, 1997 (available online:

https://www.researchgate.

net/publication/210198488_Design_and_Evaluation_for_

Situation_Awareness_Enhancement,04.07.2024).

[7] Gerigk, M. Interference between Land and Sea Logistics

Systems. Multifunctional Building System Design

Towards Autonomous Integrated Transport

Infrastructure. TransNav The International Journal on

Marine Navigation and Safety of Sea Transportation, 16,

439-446, 2022. (https://doi. org/ 10.12716/1001.16.03.04).

[8] Goodwin E. M. A Statistical Study of Ship Domains The

Journal of Navigation , Vol. 28, 3 , pp. 328 – 344, 1975.

(doi: https://doi.org/10.1017/S03734633000412 30).

[9] Hansen, M. G., Jensen, T. K., Lehn-Schiøler, T. , Melchild,

K., Rasmussen, F. M. and Ennemark F. Empirical Ship

Domain based on AIS data. Journal of Navigation 66.6,

pp. 931–940, 2013 (doi::10.1017/s0373463313000489).

[10] Kahneman D., Tversky A., Prospect theory: An

Analysis of Decision under Risk. Econometria, 47(2), pp.

263-291. March 1979 (available online

https://www.jstor.org/stable/1914185 on 04.07.2024).

[11] Muller-Plath G., Lehleitner J., Maier J., Silva-Löbling J.,

Zhang H., Zhang X., Zhou S. How Does Maritime

Situation Awareness Depend on Navigation Automation

and Mental Workload? A Sea Simulator Experiment.

TransNav The International Journal on Marine

Navigation and Safety of Sea Transportation,Vol. 17, No

4, 2023. DOI: 10.12716/1001.17.04.23).

[12] Ozturk U. Data-driven Ship Domain for Open Water

Navigation. Journal of ETA Maritime Science 10(1), pp.

39-46, 2022.

[13] Pietrzykowski Z, Magaj J. Analysis of ship domains in

traffic separation schemes. Scientific Journals of the

Maritime University of Szczecin, 48 (120), pp. 88–95,

2016 (doi: 10.17402/181).

[14] Porathe T. Prison J., Man Y. Situation awareness in

remote control centres for unmanned ships. Proceedings

of the Conference: Human Factors in Ship Design &

Operation. London, UK, 2014,

(doi:10.3940/rina.hf.2014.12).

[15] Testing of RBAT on specific cases of MASS concepts.

DNV – Report No. 2022-0481, 2022 (available online:

www.dnv.com, 04.07.2024).

[16] Thombre S. , Zhao Z., Ramm-Schmidt H., García J. M.

V., Malkamäki T., Nikolskiy S., Hammarberg T., Nuortie

H., Zahidul M., Bhuiyan H., Särkkä S., Lehtola V. V.

Sensors and AI Techniques for Situational Awareness in

Autonomous Ships: A Review IEEE Transactions on

intelligent transportation systems, Vol. 23, 1, 2022.

[17] Weintrit A. Time to Revise the IMO’s Guidance on

Good Practice for the Use of Electronic Chart Display

and Information System (ECDIS). TransNav, the

International Journal on Marine Navigation and Safety

of Sea Transportation, Vol. 16, No. 3, pp. 523-531, 2022

(doi:10.12716/1001.16.03.15).

[18] Weintrit A. Revision of the IMO’s Performance

Standards for ECDIS. Three Versions of Performance

Standards in Use. TransNav, the International Journal on

Marine Navigation and Safety of Sea Transportation 16,

4, pp. 675-683, 2022 (doi: 10.12716/1001.16.04.09).

[19] Wrobel K. A Tale of Two Disruptive Maritime

Technologies: Nuclear Propulsion and Autonomy.

TransNav, the International Journal on Marine

Navigation and Safety of Sea Transportation 16, 4, pp.

733-741, 2022 (doi: 10.12716/1001.16.04.15).

[20] https://www.imo.org/en/MediaCentre/HotTopics/Pages

/Autonomous-shipping.aspx (available online,

04.07.2024).

[21] https://seaperformer.com/ (available online,04.07. 2024).

[22] https://www.sofarocean.com/ (available online

04.07.2024).