International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 3

Number 1

March 2009

Development of a Concept for Bridge Alert

Management

F. Motz & S. Hockel

Research Institute for Communication, Information Processing and Ergonomics, FGAN,

Wachtberg, Germany

M. Baldauf

World Maritime University, Malmö, Sweden

K. Benedict

Hochschule Wismar University of Technology, Business and Design, Warnemünde,

Germany

1 INTRODUCTION

Modern ship bridges show a high degree of automa-

tion. A large amount of information concerning nav-

igation, safety and security as well as the monitoring

and control of the technical facilities on board are in-

tegrated in the operational displays on the bridge.

With respect to the level of integration of the sen-

sors, equipment, displays and assistance systems a

modern ship bridge can be defined as a highly-

complex man-machine system. As such the safety

and efficiency of its handling is dependent on the

communication between the human and the ma-

chines during the accomplishment of tasks. Humans

can fulfill their assigned monitoring, control, and

decision tasks most effectively, if the information

flow between them and machines is adapted to the

human skills and abilities (e.g., Lützhöft 2004). To

support the mariner effectively, information should

be presented task- and situation-dependent.

Associated with the high degree of automation

and integration of systems and sensors on board is a

proliferation of alarm signals on the bridge. Redun-

dant and superfluous audible and visual alarms are

appearing, without a central position for visualiza-

tion and acknowledgement of alarms. That way

alarms increase the seafarers’ workload and lead to

information overload (Earthy 2006). Alarm signals

coming from various systems and sensors lead

sometimes to a confusing and difficult manageable

situation for the mariner, which is distracting him

from his task to safely navigate the vessel.

The majority of marine accidents are associated

with collisions and groundings – in terms of num-

bers as well as in terms of costs. While some acci-

dents result from technical failures of one kind or

another (e.g. structural, engine or steering failure),

many are caused by navigational errors (Wadsworth

2005). Research indicates that a high percentage of

collisions and groundings are due to direct human

error (IMO 2008).

To enable the operator to devote his full attention

to the safe navigation of the ship and to immediately

identify any abnormal situation requiring action to

maintain the safe navigation of the ship an alarm

management harmonizing the handling, distribution

and presentation of alarms on the bridge is necessary

(Bainbridge 1983, Sheridan 1998).

ABSTRACT: Modern ship bridges are highly-automated man-machine systems. Safe and efficient ship op-

erations are dependent on the communication between humans and machines. This paper is dedicated to the

general subject of integrated navigation and the specific field of the alert management on a ship's navigational

bridge. It deals with investigations into the present situation on board of ships regarding the frequency and

type of triggered alarms under real conditions. The conduction of empirical field studies is introduced and

some of the gained results are presented and discussed. Finally the alert management concept of the perfor-

mance standards for Integrated Navigation Systems (INS) is introduced and an approach for the reduction of

CPA/TCPA alarm frequencies within INS/IBS is described.

For navigational alarms such an alarm manage-

ment is introduced within the revised performance

standards for integrated navigation systems (INS),

which needs to be implemented for all INS installed

after January 1

2011. Additionally, the Internation-

al Maritime Organization (IMO) decided to develop

in the context of the revision of the performance

standards for integrated bridge systems (IBS) a

bridge alarm management system that comprises all

alarms occurring on the bridge. A correspondence

group coordinated by Germany was established to

progress this work.

The importance of an alert management is recog-

nized as well within the framework of the e-

navigation strategy of the IMO (IMO 2008). As one

high-level user need within the e-navigation strategy

an alarm management system as accomplished in the

revised performance standards for INS is identified.

To investigate the current situation of the man-

agement and presentation of all alarms appearing on

ships’ bridges a number of field studies were per-

formed on board of vessels. The results of those

studies were introduced in the work of the IMO cor-

respondence group working on the development of

performance standards for Bridge Alarm Manage-

ment. Within this paper selected results of the inves-

tigations are presented as well as the conclusions

drawn regarding the performance standards for

Bridge Alarm Management which will be finalized

at the 55

session of the IMO Subcommittee Safety

of Navigation (NAV) in 2009. An approach for the

reduction of the number of alarms is introduced.

The field studies were performed under the

framework of a national Research and Development

project funded by the German Ministry of Transport,

Building and Urban Affairs, and under the European

MarNIS-project, funded by the European Commis-

sion, Department for Energy and Transport.

2 FIELD STUDIES – METHOD AND SAMPLE

Field studies were carried out on board of six ves-

sels: two ferries operating in the Baltic Sea, three

container vessels (with container capacities of 6200

TEU, 5500 TEU and 7500 TEU) and a cruise vessel

operating in the Mediterranean Sea. The aim of the

field studies was to investigate the current occur-

rence of alarms on a ships bridge and their handling

by the bridge team. As the management and presen-

tation of alarms is influenced by the type of ship, the

year of construction, the installed equipment and

grade of integration, the sea area, the training and

education of the crew as well as by the safety stand-

ards of the shipping company (Baldauf & Motz

2006), these factors were taken into account to ob-

tain a profound database.

All alarm and warning messages occurring on the

bridge were manually recorded. It was registered

what kind of alarm occurred, when and where it was

announced, how it was presented and how it was

handled by the bridge team.

To compare the different bridges on each vessel

the specific sensor systems in use for navigation

(positioning, speed, track control, collision avoid-

ance and so on) together with the configuration of

the alarm thresholds were registered. Simple chang-

es in the configuration and the settings of the alarm

limits lead to an increase or decrease of the an-

nounced alarms.

A special focus was laid on the assumed depend-

encies of alarm frequencies from sea areas. For the

purposes of the studies the navigational situations

were defined based on collected experts’ opinions

as:

− "open sea" (no natural constraints / no artificial

constraints),

− "coastal" (natural constraints / distance to coast

less than 10 nm without harbor, pilotage, anchor-

age / artificial constraints - e.g. TSS with estab-

lished traffic lanes and recommended routes) and

− "confined" (with three different special cases:

harbor, pilotage and anchorage).

Additionally interviews by means of a structured

guideline were carried out to gather opinions, sug-

gestions and remarks of navigators on the occur-

rence and presentation of alarms on the bridge, the

handling of alarms and related operational problems.

The investigated vessels were built or recon-

structed within the time span from 2001 until 2007.

The ships’ navigational bridges were provided with

equipment from different manufacturers that was in-

tegrated and combined on a medium or high integra-

tion level.

The field studies were conducted on different

times of the year during voyages in the Baltic Sea, in

the Western Mediterranean Sea, in the North Sea

and in the English Channel. Usually good weather

conditions were experienced with low winds and

calm sea. The average time of observation was 19

hours, with a minimum of 11 hours and a maximum

of 27 hours.

3 FIELD STUDIES – RESULTS

3.1 Alarm thresholds and adjustment of alarm

announcement

In the framework of the field studies it was recorded

which settings the bridge team used for the trigger-

ing of alarms and warnings at the installed naviga-

tional equipment. It was assumed that the navigators

would adjust the alarm thresholds and limit values to

the different navigational situations and sea areas.

However the contrary was observed. On three of

the six vessels no adjustments at all were made to

the alarm thresholds and limit values – neither ac-

cording to sea areas nor visibility conditions. On the

remaining ships a change of the thresholds was rare-

ly the case as well. One of the navigators adjusted

the limits for CPA/TCPA (Closest Point of Ap-

proach/Time to CPA) in confined conditions to 0.5

nautical miles (nm) and 0 minutes (min) to practical-

ly switch off the alarms.

In general the thresholds for CPA alarms were set

to 0.5 nm on five of the vessels with a TCPA limit of

6, 10, 12 or (twice) 15 minutes. On Ferry 2 the

CPA/TCPA limits were set to 0.0nm/0min through-

out the voyage.

Adjustments to the different navigational situa-

tions and sea areas were observed regarding the

alarm announcement. The audible signaling of radar

alarms was switched off from time to time on almost

all vessels. Especially in confined conditions during

departure and arrival, where a lot of targets appear

on the radar, alarms (e.g. collision avoidance and

lost target alarms) were not acoustically announced.

On Ferry 2 the audible alarm announcement was

switched off on both radars throughout the voyage.

3.2 Dependencies on sea area

One hypothesis of the investigations was that the

frequency of alarms is dependent on the sea area in

which the vessel is sailing. Figure 1 shows the aver-

age frequency of alarms per hour for the three sea

areas for each vessel investigated.

0,0

5,0

10,0

15,0

20,0

25,0

30,0

Ferry 1 Container

vessel 1

Ferry 2 Container

vessel 2

Cruise

vessel

Container

vessel 3

Frequency of alarms per hour

open sea

coastal

confined

Figure 1. Average frequency of alarms for sea areas per vessel

Except for Ferry 2, on all vessels considerably

more alarms were recorded in coastal and confined

conditions than in open sea areas. Altogether the

analyses indicate a correlation between the traffic

characteristics of the specific sea areas and the alarm

occurrence.

This hypothesis is further confirmed when ana-

lyzing the average frequency for all six vessels (Fig.

2). Most alarms were recorded in confined and

coastal waters. The occurrence of alarms at open sea

was approximately four times lower than in confined

waters.

4,4

11,8

16,2

0,0

5,0

10,0

15,0

20,0

25,0

30,0

open sea coastal confined

Frequency of alarms per hour

Figure 2. Average frequency of alarms for sea areas for all six

vessels

3.3 Dependencies on equipment

Figure 3 depicts the distribution of alarms dependent

on the device on which they occurred for the six

vessels.

The vast majority of alarms was recorded at the

radar device. Only on Container vessel 3 the per-

centage of radar alarms was lower than the amount

of alarms that occurred at the ECDIS. This is due to

the fact that on this vessel throughout the voyage

AIS information was not integrated in the radar. Al-

together the ECDIS aggregates the second highest

percentage of the registered alarms.

The ECDIS was the only system besides the radar

for which on each of the vessels alarms were record-

ed. Alarms on other devices were not appearing on

all vessels, in some cases only on one vessel. These

alarms reflected specific circumstances on the con-

cerning vessel. On Container vessel 2 for example

many alarms were registered on the GNSS and the

gyro monitoring system, caused by the loss of the

differential signal at the GNSS device.

20%

40%

60%

80%

100%

Ferry 1 Container

vessel 1

Ferry 2 Container

vessel 2

Cruise

vessel

Container

vessel 3

Percentage of alarm occurence

Radar

ECDIS

Other

Figure 3: Distribution of alarms dependent on equipment per

vessel

3.4 Dependencies on alarm category

The distribution of the five alarm categories regis-

tered most often is shown in Figure 4.

For the purpose of the analysis the percentages

for off track and off course alarms were summed up

into one category. The category “collision avoidance

alarms” includes the values for CPA/TCPA and Bow

Crossing alarms. Aggregated into the category

“waypoint alarms” are the alarms: early course

change indication, actual course change indication

and wheel over point. Alarm types summed up under

the category “other” partly took percentages of up to

17% on one vessel but were registered only on that

vessel – as for example DGPS failures on Container

vessel 2.

Other

23%

Waypoint alarms

Off track / Off

course

12%

Lost target

20%

Chart data

warning

Collision

avoidance alarms

31%

Figure 4: Average percentage of alarm categories for all six

vessels

For all vessels investigated the majorities of

alarms are collision avoidance alarms and lost target

alarms (51 %).

CPA/TCPA as well as lost target alarms were

predominantly triggered by AIS information, on av-

erage 72% of the CPA/TCPA alarms and 57 % of

the lost target alarms. (This percentage could have

been even higher, if the bridge team of Container

vessel 3 had not chosen a radar setting without inte-

gration of AIS information, which caused that all

CPA/TCPA and lost target alarms were initiated by

radar information.)

These results were expected due to the technical

configuration and the use of the automatic alarm

functions. For AIS, according to IMO regulations,

the same limit values have to be applied as for

tracked radar targets and the option for CPA/TCPA

calculation was switched on to sleeping AIS targets

by default. On two ships it was observed that

CPA/TCPA and lost target alarms included targets

that lay in the harbor basin behind land masses.

3.5 Interview results

During the field studies on board 13 mariners were

interviewed. All of them were male with an average

age of 36. Their average overall experience as mari-

ner was 14 years, their average experience as an of-

ficer 9.5 years. The current position of the inter-

viewees varied from master to third officer – most of

the mariners were first or second officer.

Generally officers and masters feel that there are

too many alarms occurring only for informative rea-

sons on the bridge or can not be solved by the officer

on the watch, for example “VDR record fault” or

“window wiper oil low”. Half of the mariners think

that especially in certain situations for example the

approach to a harbor such alarms are a major prob-

lem as they are distracting.

Mariners often reported alarms making annoying,

loud and long lasting sounds, for example alarms

from the navigation lights, echo sounder or gyro

compass. Others pointed out that the acoustic

presentation of an alarm often doesn’t reflect its rel-

evance. To identify the priority and the source of an

alarm is seen as a general problem.

The majority favors to have the possibility to

switch of the audible announcement of alarms in

certain situations. Some mariners say that this con-

cerns especially noisy alarms that are seen as less

important. Others are referring to situations in which

there is a lot of traffic, instruments are closely moni-

tored and alarms are expected. However it is stated

by some mariners that alarms for safety reasons, like

fire alarms or engine shut down alarms, should al-

ways be audible.

Eighty percent think that a centralized alarm dis-

play for the centralized presentation of all alarms

would support them with their tasks on the bridge.

The vast majority feels that it would be a great bene-

fit not to have to “run around the bridge” anymore

trying to find out from which equipment an alarm

comes from. Further expectations regarding a central

alarm management display include a prioritization of

alarms and the possibility to acknowledge or at least

silence the alarms there. Some mariners worry that

safety related alarms will not be immediately identi-

fied at a central alarm display for all alarms.

False alarms are seen as a problem. According to

mariners especially distress alarms are often false

alarms caused by users that send a message by mis-

take (without identification or position). Another

problem are distress alarms from areas which are not

of interest for the actual navigation situation.

Regarding the handling of alarms from the engine

room presented e.g. on extension control panels on

the bridge the statements differed. On some vessels

the engine room is manned all or nearly all the time

and alarms from the engine room are presented only

for informative reasons without giving an audible

announcement. On those vessels no difficulties with

engine alarms are experienced. On other vessels

alarms from the engine room give a sound on the

bridge that stops, when the alarm is acknowledged in

the engine room. On the bridge these alarms can not

be acknowledged. It is only possible to silence the

audible alarm. The tone is retriggered, when a new

alarm occurs. This can lead to a distraction of the

navigator, whose attention gets attracted by the

sound again and again. Some mariners feel that

those engine alarms are annoying.

4 DISCUSSION OF RESULTS

The field studies indicate a lack of a harmonized

alarm management. The majorities of alarms record-

ed were collision avoidance and lost target alarms

occurring on the radar, being triggered by AIS in-

formation. The peak values of alarms per hour were

observed in confined waters. Under conditions of

high traffic density, as for example at harbor en-

trances, alarms were often experienced as superflu-

ous. However alarm thresholds were rarely adjusted.

Instead the audible alarm announcement on the radar

often was switched off in confined conditions. On

some vessels the bridge teams suppressed collision

avoidance alarms during departure and arrival by

setting the alarm thresholds to a minimum. Alto-

gether it can be concluded that especially in con-

fined waters as harbor approaches where many AIS

targets appear the navigators are overloaded with

alarms.

The results showed further, that difficulties are

related to the audible presentation of alarms and the

necessary acknowledgement on various panels on

the bridge, to the lack of indication of any priority of

the alarms, to the lack of a consistent alarm ac-

knowledgment concept and to difficulties in differ-

entiating the audible alarm signals.

4.1 Generic Approach for Reduction of CPA/TCPA

alarms

As collision warnings were found a major part of all

alarms on the bridge the situation regarding high

number of alarms may possibly be improved by en-

hanced triggering of these type of alarm.

One reason for the high amount of collision

avoidance and lost target alarms in confined waters

is to be seen in the technical configuration of AIS

and the integration of sleeping AIS targets for the

presentation of collision avoidance and lost target

alarms. According to the new display and the new

radar standards (IMO 2004a, IMO 2004b), future ra-

dar systems with AIS integration will allow the se-

lective acquisition of AIS targets for collision avoid-

ance alarms and more flexibility for the presentation

of lost target alarms. That way future radar system

will have the capability to reduce the number of

alarms. However, further studies are needed, to in-

vestigate if this really will solve the problem.

Further reasons for the high amount of alarms in

confined waters are on the one hand the missing of

recommendations for thresholds to be used for CPA

and Bow Crossing limits taking into account rele-

vant situation parameters and on the other hand the

missing possibilities to adjust the alarm initiation in

an appropriate way beyond range and time to con-

fined waters, where closer passing of vessels are to

be expected.

One concept for the improvement of collision

warnings has already been described earlier by Bal-

dauf (2004). It bases on the definition of situation

dependent thresholds, which take into account the

type of encounter situation, the sea area and the visi-

bility conditions. Core element of this approach is a

risk model for situation assessment defining the

three types of encounters – meeting, overtaking and

crossing courses - and considering the two condi-

tions of good and restricted visibility as laid down in

the International Rules for Preventing Collisions at

Sea. To reduce the number of collision warnings the

situation-dependent thresholds for CPA and TCPA

can be applied by an algorithm for self-adaptation of

these values to the prevailing circumstances of a cer-

tain situation and the maneuvering characteristics of

the involved ships. According to first preliminary

tests using a set of determined initial situation-

dependent CPA values to recorded open sea scenari-

os a significant reduction of occurring collision

warnings was reached (Baldauf et al. 2008).

5 PERFORMANCE STANDARDS FOR BRIDGE

ALERT MANAGEMENT

A lot of deficiencies observed in the field studies

will be solved for INS installed after January 1

2011 for which an alarm management, according to

the revised INS performance standards (IMO 2007),

is mandatory.

This alarm management system for navigational

alarms aims to harmonize the operation, handling,

distribution and presentation of alarms. To improve

the operator’s situation awareness and his ability to

take effective action a set of priorities is introduced

based on urgency of the required response. A new

philosophy is followed for the prioritization and cat-

egorization of alarms. Alert (alert management) is

defined as umbrella term for the indication of any

abnormal situation with three different priorities of

alerts (IMO 2007, Motz & Baldauf 2007):

− alarm (highest priority) - conditions requiring

immediate attention and action by the bridge team

to avoid any kind of hazardous situation and to

maintain the safe navigation of the ship;

− warning - conditions or situations which require

immediate attention for precautionary reasons, to

make the bridge team aware of conditions which

are not immediately hazardous, but may become

so; and

− caution - awareness of condition which does not

warrant an alarm or warning condition, but still

requires attention out of the ordinary considera-

tion of the situation or of given information.

The three priorities are indicated visually and acous-

tically in different ways. Whereas alarms initiate an

audible signal and a flashing visual indication until

acknowledgement, warnings are presented with a

momentarily audible signal and a flashing visual in-

dication until acknowledgment. After acknowledg-

ment both alarms and warnings are presented with a

steady visual indication. Cautions are only indicated

by a steady visual indication and don’t have to be

acknowledged. It is also possible to temporarily si-

lence alarms.

To ensure a consistent presentation of alerts and

to reduce the presentation of high priority alerts

within the INS, alerts released by navigational func-

tions, sensors, sources are presented as far as practi-

cable, after evaluation with the system knowledge of

the INS, e.g., provided by the integrity monitoring.

This means that the priority of an alert will be as-

signed and presented consistently for all parts of the

INS, and can be reduced for the alert in case of suf-

ficient redundancy. E.g., in case of a failure of one

of three position sensors only a caution may be re-

leased for the INS as still a reliable system position

can be presented.

Additionally the INS performance standards in-

clude requirements for a central alert management

human machine interface (HMI) for navigation re-

lated alerts (IMO 2007). Such a centralized presenta-

tion is part of an INS to support the bridge team in

the immediate identification of any abnormal situa-

tion, including the source and reason for the abnor-

mal situation and information for decision support

for the necessary actions.

The central alert management HMI has to fulfill

three major functions: indicating and identifying

alerts, allowing to temporarily silence all alarms and

allowing the acknowledgement of all alarms and

warnings for which no additional graphical infor-

mation is necessary as decision support for the eval-

uation of the alert related condition.

The alert management system within INS was

developed with the intention to be extendable to an

alert management concept for the whole bridge. The

findings of the field studies showed that the aspects

which were contributing to the development of the

INS alert management are also to be applied to the

alert management system for all alerts on the bridge.

Accordingly the performance standards for

Bridge Alert Management, which is currently devel-

oped at IMO, picks up most of the ideas of the INS

alert management. In doing so the performance

standards consists of two major parts: A general

module aiming to harmonize the presentation and

handling of all alerts on the bridge (equivalent to the

prioritization introduced within INS) and additional-

ly requirements for a central alert management HMI.

ACKNOWLEDGEMENTS

The field studies were part of a project funded by the

German Ministry of Transport, Building, and Urban

Affairs, and conducted under the framework of the

European MarNIS-project, funded by the European

Commission, Department for Energy and Transport.

The authors would like to thank the shipping

companies Peter Döhle, TT-Lines, Finnlines,

Scandlines, HAPAG-Lloyd and AIDA Cruises Ltd

for their grateful assistance and all mariners who

provided their knowledge in interviews on board.

REFERENCES

Bainbridge, L. 1983. Ironies of Automation. Automatica 19:

775-779.

Baldauf, M. 2004. Enhanced Warning Functions for on board

Collision Avoidance using AIS and VDR information. In R.

Dauer, A. de la Pena, J. Puig (eds), International Congress

on Maritime Technological Innovations and Research-

Proceedings. Barcelona.

Baldauf, M.; Benedict, K. & Motz, F. 2008. Aspects of Tech-

nical Reliability of Navigation Systems and Human Ele-

ment in Case of Collision Avoidance. Proceedings of the

Navigation Conference & Exhibition, London, UK.

Baldauf, M. & Motz, F. 2006. Operational Aspects of future

Alert Management to support ship navigation. (in German)

In Morten Grandt (ed.), Cognitive Systems Engineering in der

Fahrzeug- und Prozessführung. Bonn: DGLR-Bericht 2006-02.

Earthy, J. 2006. Raising the alarm. Horizons 15: 10-11.

IMO 2004a. Performance standards for the presentation of nav-

igation-related information on ship borne navigational dis-

plays. MSC.191(79). London: International Maritime Or-

ganization.

IMO 2004b. Revised performance standards for radar equip-

ment. MSC.192(79). London: International Maritime Or-

ganization.

IMO 2007. Revised performance standards for integrated navi-

gation systems (INS). MSC.252(83). London: International

Maritime Organization.

IMO 2008. Report of the Subcommittee on Safety of Naviga-

tion to the Maritime Safety Committee. NAV 54/25. Lon-

don: International Maritime Organization.

Lepsoe, A. & Eide, M. 2005. Field Study on Bridge Alarms.

Interim Sub-Task Report MarNIS WP2.4. Oslo: Det Norske

Veritas.

Lützhöft, M. 2004. Maritime Technology and Human Integra-

tion on the Ship's bridge. Dissertation No. 907, Linköpping

Studies in Science and Technology. Sweden.

Motz, F. & Baldauf, M. 2007. Investigations into Shipborne

Alarm Management - Conduction and Results of Field

Studies. Proceedings of the 9th International Conference on Enterprise In-

formation Systems, Volume HCI: 136-141. Funchal, Portugal.

Sherwood Jones, B.M.; Earthy, J.V.; Fort, E.; Gould, D. 2006.

Improving the design and management of alarm systems.

Proceedings of the World Maritime Technology Confer-

ence, London, March 2006.

Sheridan, T.B. 1998. Rumination on Automation. In S.

Hishida; K. Inoue (ed.), Analysis, Design and Evaluation of

Man-Machine-Systems. Reprints of the 7th

IFAC/IFIP/IFORS/IEA Symposium, Kyoto.

Wadsworth, B. 2005. Marine eNavigation: An orientation pa-

per. 9th world-wide electronic navigational chart database

committee. WEND 9/INF.4. Monaco: International Hydro-

graphic Organization