INS and IBS Problems and Incidents

To record your problems in the comments box, it is necessary to be a registered user. This gives you a user name (Wikiname) and you give it your password.

Registration is at http://www.maresig.org/twiki/view/TWiki/TWikiRegistration (also a direct link from the http://www.maresig.org home page)

There is a lot of good material at http://www.panbo.com/ which needs collating if someone has the time.

Problems

From MSA New Zealand report on Aratere near grounding:

The following is an extract from a paper prepared by Rothblum, Sanquist, Lee and McCallum. “Identifying the effects of shipboard automation on mariner qualification and training and equipment design. Paper prepared ISHFOB 95: Informational Symposium – Human Factors on board. Bremen, Germany. November 15-17, 1995.

“Automation is becoming more prevalent on commercial ships, affecting such areas as engineering, bridge and cargo operations. When designed properly and used by trained personnel, such automation can be helpful in improving operational efficiency and safety. However, when designed poorly or misused by under-trained or untrained personnel, automated equipment can be a contributing cause to accidents. In one study of 100 marine casualties, inadequate knowledge about equipment was found to be a contributing cause in 35 percent of the casualties. The most frequently cited problem was the misuse or non-use of radar. “Lack of training is not the only problem. Poor equipment design can induce the mariner to make mistakes. In the same study, 1/3 of the accidents were found to be caused partly by poor human factors design of the equipment.”

Equipment from different manufacturers is different; one way of reducing the impact of this is to have a consistent structure for manuals so that at least the user knows where to look things up. Suggestion from Mike Powell, CHIRP.

-- TWikiGuest - 28 Jan 2006

MAIB Bridge Watchkeeping Safety Study Safety Study 1/2004 (2004)

“As technology has advanced with regard to radar and ARPA, and the number of crew on vessels has decreased in parallel with increased automation, it is inevitable that bridge watchkeeping practices have changed in recent years. OOWs place more reliance on radar and ARPA to maintain a lookout, and to assess the risk of collision. Indeed, many newer vessels are not even equipped with a gyro pelorus on the bridge with which to take a visual bearing.”

Lützhöft Dissertation (2004)

Examples include:

p64 Cross-track error shown as Left and Right (with very small numbers).

p74 MOB chart system that requires th operator to go through 5 steps to register the position (submenus, button pushes) at the same time as he has to start turning the ship, call the captain and crew, sound alarms and launch an MOB lifebuoy. Other systems allow this with only one push of a button.

p74 When a radar that was part of an integrated navigation system had to be switched off, the crew did not know which of the other parts would ‘stay on’.

Paper 1 - Integration work on the bridge (Lützhöft and Nyce)

p29 Loss of GPS signal not indicated on ECDIS – ECDIS does not reflect loss of signal.

Paper 4 Studying the effects of technological change

p111 in course mode, under certain conditions such as when reducing the ship’s speed, the course corrections are not calculated and implemented rapidly enough.

‘Narrow margins leave no room for error’ The Swedish Club Letter Number 2, November 2005

http://www.swedishclub.com/lm_publications/2005-2.pdf

"The widespread adoption of new collision avoidance technology has had no discernable effect on the high incidence of collisions. Following a new study of major collisions, the Club now warns that the new technology can promote over-confidence and complacency.

...

One issue already identified is the concentration of vessels on the ‘optimum track’ as indicated by the GPS, but there are also other factors at work.

...

It will require further study of a larger ship population to prove the point beyond contention, but we already have evidence that the new technology presents new hazards.

...

There is a temptation to accept narrow margins which leave no room for error. There is then the danger that vessels in close proximity fail to manoeuvre as expected. We are moving from the well-documented problem of ‘radar-assisted collisions’ to a new phenomenon: deviation from correct procedures simply because an electronic display suggests that a manoeuvre should be safe. Safety margins exist to accommodate the possibility of error or misunderstanding. People need to be fully aware that ‘having the full picture’ is not a licence for dispensing with the International Collision Regulations and the fundamentals of good seamanship."

CHIRP Parallel Index Problems

The difficulty of setting up and adjusting Parallel Index lines with many modern systems has been the subject of an investigation reported to CHIRP about which the CHIRP Maritime Advisory Board were of the view that interested bodies should be informed. See letter to MarESIG and report below.

-- BrianSherwoodJones - 23 Dec 2005

• CHIRP_parallelindexletter.pdf: Letter to MarESIG^? on problems of Parallel Index Lines .pdf is 1.7MB.

• CHIRP_PI_examples.pdf: Steps required to set up Parallel Index (PI) Lines and adjust them. .pdf is 1.1 MB.

INTEGRATED BRIDGE AND NAVIGATION SYSTEMS USER ENHANCED DESIGN

Official report form The Nautical Institute’s IBS/INS Conference 13-14 November 2002, London

http://www.nautinst.org/pdf/IBS%20Conf%20Report%20Final.pdf

THE TECHNOLOGY AFFECTING MODERN BRIDGE SYSTEMS

(Captain Chris Cobley, FRIN)

CIRM acknowledges concerns about the problems of managing the ever-increasing amounts and types of navigational information now becoming available to the mariner, and in particular how this can be correlated and presented in the most useful, clear and unambiguous way. They assert that unless the available information from Radar, ECDIS and AIS is properly correlated and presented in a way that can be quickly and unmistakeably assimilated, there is a real risk that more information may overload rather than assist. For example:

• Existing Radar and ECDIS systems have their own separate graphical displays, each with its own different symbology. This makes it difficult for the user to assimilate instinctively, and increases the likelihood of mistakes.

• A radar image superimposed on ECDIS can be very clear and easy for the mind to correlate, but anything more complicated than this can start to overload, not only the mind of the user, but also the displays themselves.

• The IMO requirement is for essential information received by AIS to be displayed on a Minimum Keyboard Display (MKD) - a very basic display, which will be of limited value to the watchkeeper. Current radar processors and displays cannot handle the volume of data generated by AIS. IMO SN/Circ.217 provides interim Guidelines for the display of selected information within the limited capacity of the current radar processor and display. The need to display AIS information to the OOW in a more useful, graphical format is well recognized, but how best to meet this need has been a matter of some dispute.

• There are incompatibilities between Radar and AIS symbologies, which need to be resolved in order to provide the navigator with an unambiguous and intuitive picture.

THE PILOT’S PERSPECTIVE (Captain Geoff Taylor, MNI)

Key concerns are the way in which information is presented in terms of its symbology, clarity and discrimination of immediate and short-term hazards, and a lack of commonality in design and symbology on the switches and controls of most navigation and communications equipment.

DEVELOPING INTEGRATED BRIDGE AND NAVIGATION SYSTEMS THE MANUFACTURER’S VIEWPOINT (Hubertus (Fritz) Wentzell)

Navigational information should be available with the quality and accuracy required for the task at hand. It should be unambiguously clear for any mariner trained in the understanding and use of INS, reliably and redundantly obtainable, even in case of a technical failure and easily adaptable to different nautical situations. To this end, integration of information should be capable of:

• Using all available sources.

• Automatically checking the validity and integrity of data.

• Ensuring fail-safe redundancy.

• Supporting unambiguous situation awareness.

• Automatically providing warnings and alarms for arising hazards.

• Offering an ergonomic human-machine interface.

There may be divided opinions amongst the users about the integration of navigational information. For example: • Integration of information reduces the ambiguity of information, but requires additional training to understand the process of integration. Are our mariners suitably trained on state of the art navigation equipment and in appropriate time intervals, regardless of integrated systems?

• Integration of information provides automatic warning and alarm functions, but may also lead to overconfidence of mariners. Is this not only a matter of training, but also of adequate bridge procedures?

• Integration of information reduces the number of displays and indicators to be monitored, but some mariners may want to determine themselves from which equipment the information is most useful. Could this require even more training than simply understanding an integrated system?

• Integration of information automatically checks the planned route for known hazards, i.e. dangerous chart objects, but some mariners may feel that the detailed route planning functions are too tedious. Can the marine community, and the environment, allow inadequate route planning and checking for hazards?

TRAINING IMPLICATIONS OF NEW TECHNOLOGY (Captain Gustaf Gronberg, MNI)

The information offered in User’s Manuals tends to be too extensive and technical for a navigator to comprehend. This problem can be solved by the production of a user-oriented Bridge Familiarization Manual that is unique to every ship; but, the suppliers of bridge systems must take a higher level of responsibility for the navigator’s understanding of complicated systems, by supplying the ship with a computer based training program which focuses on the parts that are essential to the navigator.

Discussion

There is a recognition that in reality, although equipments are type approved and IMO compliant, to minimum performance standards, the manufacturers add their own features, in an attempt to make their equipment ‘user-friendly’. For example, the mariner can either be faced with joystick, trackball or menu-driven controls, depending on the equipment fit in the vessel in which he is serving. This can cause him to be distracted from the primary task of watchkeeping, which is to conduct and monitor the safe navigation of the ship, including the keeping of a good lookout.

It was noted that there is now a plethora of non-standard - albeit type approved - systems, both stand-alone and integrated, where training is not yet an integral part of the introduction into service of that new technology and equipment. Not surprisingly, all the participants agreed that the watchkeeper needs to be trained on new technology and equipment; he should not be expected to pick it up after he has joined the vessel, or to undergo familiarisation by other staff onboard, who themselves have no formal training or qualifications in the use of such equipment. Key questions were whether it is practicable for mariners to undertake type specific training for each different manufacturer’s system and whether it should be part of formal training onboard a ship before any person can take a watch.

There is some concern about the plethora of alarms (in one of the presentations it was suggested that there could be as many as 200) that the watchkeeper may have to cope with during the course of his watch. They are perceived not only an irritant but also as dangerous in that the watchkeeper can be distracted from his navigational duties (particularly keeping a good lookout) while he is attending to an alarm. It was noted that Lloyds Classification, in their IBS notation, have taken this onboard and there are a number of ships where alarms are brought into a central location which are prioritised, with different colours and sounds depending on the nature and urgency of the warning. While this was welcomed, it was further noted that although there is the facility for single point acknowledgement, there is, as yet, no standardisation, eg Navtex may not have a remote alarm facility.

Some mariners are not readily able to recognise the changes in position of aids to navigation, or the application of Temporary and Preliminary (T&P) Notices and radio-navigation warnings, on electronic charts in the same way that they are able to with paper charts. It was pointed out that when the mariner does it by hand, he knows what the correction is, because he has personally applied it to the chart. But, if all corrections were applied automatically overnight, there is a possibility that the mariner would not notice there had been changes. It was noted that a specific requirement in the ECDIS performance standard states that ECDIS should allow the mariner to display updates in order to review their contents and ascertain that they have been included. It was explained by the manufacturers that the chart additions facility in ECDIS is a difficult item to build into the system, such that there are complications in getting user input into the system. The importance of incorporating all relevant chart information - including the movement of aids to navigation, T&P notices and radio-navigation warnings - into the Passage Plan, was emphasised.

Should there be separate displays for each function or should they be integrated into one display? It became clear that the specific needs of the deep-sea watchkeeper, the coastal navigator and the pilot may be different, depending on the operational environment.

Opinion was divided, as to the best solution. It was evident that one of the benefits of integrating information into the one display is the flexibility afforded to the user depending on his specific task within the bridge management team, eg the master, the watchkeeper, the pilot. But, the fusion, matching and correlation of data and the layering of the information is critical. The view of the mariner was that while integrated systems may be the way ahead, there is a need to maintain the flexibility to switch to stand-alone displays of, say, the ECDIS and radar pictures. To this end, it was agreed that the minimum number of displays stipulated in the current carriage requirements should not be reduced.

There was much evidence to suggest that some mariners fix their position by GPS, onto the electronic chart, and then cross check it against the second GPS. Yet, GPS is known to be vulnerable in terms of its integrity, availability and accuracy. This begged the questions:

• Should a receiver for a terrestrial radio-navigation system (eg, Loran-C or Chayka) be carried as a secondary means of position fixing, rather than as an alternative to GPS?

• Is the modern mariner sufficiently trained to be mindful of alternative methods of position fixing in the event of periods of radio and/or radar silence during hostilities, and how can this be incorporated into an integrated system?

The Conference was introduced to the concept of predictive interpretation of the Colregs on the radar display. This manufacturer-driven concept which may be well intentioned, as a means of prompting the watchkeeper of the need to take avoiding action in accordance with one or other of the Colregs, would be abhorrent to the more experienced mariner. It was felt that this was a technology-based problem-solving tool, which would ultimately cause the watchkeeper to think less for himself and further confine his awareness of the situation around him to the display rather than looking out of the window.

Integrated Bridge Systems and the Human Element The Nautical Institute held its second international conference on Integrated Bridge Systems and the Human Element on 16-17 September 2003.

http://www.nautinst.org/pdf/IBS2003Report.pdf

Much of the discussion mirrored what was said at the first IBS Conference, held in November 2002 (http://www.nautinst.org/ibs/index_2002reports.htm), particularly in relation to:

• the plethora of alarms and the potential for distracting the officer of the watch from his main purpose of conducting the safe navigation of the ship;

• the inadequacy of the pilot’s conning position;

• the need for standardisation in terms of switches and control keys;

• issues of illumination levels from equipment during periods of darkness;

• the technical vulnerabilities of the various stand-alone and integrated systems;

• generic and type-specific training in the use of both stand-alone and integrated systems.

It is, however, a very common problem that the actions of the master are not properly monitored although it forms the very basis of Bridge Team Management Training. On large bridges, there may be difficulties with the master shouting from the bridge wing; bridges have become bigger, and often the master or the chief officer has to shout from the bridge wing and perhaps the important knowledge that he is imparting has to be relayed over the noise of the surrounding environment.

There are many systems around that are not ‘integrated’ in the true sense, notwithstanding IMO and other requirements. One of the fundamental problems in the background to a recent investigation was that of two ARPAs that integrated into the system but, because they were integrated, one software fault affected them both so that they were both inoperable in terms of the collision avoidance function of the ARPA.

The IMO performance standards do not include a requirement to manually input bearings and ranges into the ECDIS - a problem that needs to be remedied by IMO when the ECDIS requirements are discussed again. However, it appears that some ECDIS have the facility to input two bearing lines and two range markers, such that in the event of primary GPS failure, position can be plotted using traditional navigation techniques.

It was reported that the use of PPUs by pilots could lead to the bridge team not properly monitoring their own navigation.

The debate concluded that: • The human tendency to lax when the pilot is onboard is probably independent of whether he has a PPU or not.

• The PPU provides another independent position fixing source allowing the diligent bridge team to cross check their own information derived from the IBS and thus ensuring that the pilot is fully integrated into the team.

• The best advice to any user of a PPU is ‘do not go anywhere with the PPU that you would not go without one’ – owing to its lack of back-up/redundancy.

• It was reported that, in the interests of best practice, trainee pilots should not be allowed to work with PPUs until they have refined their traditional piloting skills.

The manufacturers would benefit from having a set of clear international guidelines with respect to documentation and training material that they produce to assist mariners, whether it is actual onboard training or training in use of the various systems as an operational standard. It was felt that the actual specifics of it should be driven not only by the industry itself but by the IMO setting guidelines.

Crew Size and Maritime Safety

Committee on the Effect of Smaller Crews on Maritime Safety, National Research Council ISBN: 0-309-56751-3, 184 pages, 6 x 9, (1990) available from the National Academies Press at: http://www.nap.edu/catalog/1620.html

Human factors concerns have received insufficient review in shipboard automation, as they have in most complex engineering systems. Many automated systems reduce operators to passive monitors (Parasuraman, 1987; Schuffel et al., 1989) and remove much of the active content from the job without decreasing the need for vigilance. In addition, some say, the lack of standardization and the poor ergonomics of the systems make them difficult and confusing to use. One pilot told the committee, “Computerized navigational systems are designed without obtaining input from the ultimate user. They do not use common language and nautical terms to define functions. As a result, the people serving on board ship must accommodate the manufacturer and learn the specific programs involved in the equipment, rather than the other way around. At some point vessel safety will be compromised” (Bobb, 1989).

The results of experiments evaluating the impact of integrated bridge systems on bridge watch team performance have been mixed. Kristiansen, Renswick, and Mathisen (1989) found improved track-keeping and watchkeeping skills in experiments aboard seven Norwegian ships outfitted with highly automated bridges equipped with decision aids. Grabowski (1989) described the piloting expert system, a navigation aid for pilots and ship's officers. In tests at MARAD'S Computer Aided Operations Research Facility (CAORF) ship simulator, junior watch officers using the aid showed improved watch-keeping skills, but showed no significant improvement in track-keeping.

Other topics are discussed including findings from studies of one man bridge operations.

Incidents

Note that this section may duplicate items in the existing incident page, but it concentrates on the design of INS and IBS

From Risks Digest 24.33 http://catless.ncl.ac.uk/Risks

Date: Thu, 15 Jun 2006 11:47:51 -0400 (EDT) From: bo774 at freenet dot carleton dot ca (Kelly Bert Manning) Subject: Sunken Ferry Crew didn't know how to use ECS display software

Preliminary reports from the Canadian Transportation Safety board investigation into the "Queen of The North" running into Mount Gil and sinking say that the bridge crew had the Electronic Chart System Display turned off because they didn't know how to use the software control to reduce the illumination for night use.

The preliminary reports also say that bridge crew claim to not be fully aware of how to use the various steering modes, or even to know what steering mode they were in.

Digital controls should help, not hinder.

"The screen from the ECS produced too much ambient light, so the crew would often turn it off at night, Ayeko wrote. The monitor would be turned on momentarily only when it was required."

http://www.ctv.ca/servlet/ArticleNews/story/CTVNews/20060605/BC_ferry_060605/20060606?hub=Canada

This must have been an expensive system. Would it have been too much trouble to add a rotating dial or rocker button which would reduce or increase the brightness on the display? It wouldn't even have to be integrated with the monitor, just mounted somewhere close to it and clearly labeled. These don't even need to be rheostatic controls, just something that generates an input specifying the type of change requested.

Software control is bad if it makes essential functions too complex or obscure.

Some reports describe Mt. Gil as Gil Island. It is a relatively tall and steep mountain whose base is underwater. There should have been a good radar return from it. It will be interesting to see what other electronic or computer integrated safety systems also failed to make the officer and helmsman aware that their failure to change heading at the scheduled time had left them on a collision course with a mountain.

It will also be interesting to see whether the ECS brightness control issue is a "reasonable doubt" red herring raised as a defense for the criminal trial which will take place. Two passengers are missing and presumed drowned.

From Lessons Learned Summary approved at FSI 11

POLLUTION

What happened?

A tankship was properly secured to a monobuoy during cargo discharge operations. At some time during these operations, the chain stopper opened and the chafing chain was released. The ship was then moored only by a pickup rope that parted shortly thereafter. As the vessel drifted from the monobuoy, the rail hoses parted and approximately 12 tons of oil spilled into the sea.

Why did it happen?

The bridge monitor that was used to control the cargo operation used the same function keys to control different operations. The screen colour was different for each operation; however, the function keys and their sequence were not unique to a given operation. It is believed that one of the officers performing cargo operations unintentionally opened the chain stopper and released the chafing chain while attempting to secure a forward hydraulic pump. The function key sequence was the same for each operation and only the screen colour provided an indication as to which operation was being performed.

What can we learn?

Ergonomics, in the form of operator- machine interface, can be a critical element in shipboard safety.

Ship's crew should display warning signs where there is a possibility of confusion in the operator- machine interface.

From the NTSB report on Royal Majesty

"...Finally, the integrated bridge system as implemented on the Royal Majesty failed to adequately define the watch officers’ tasks and procedures. If the automation on board the Royal Majesty had been appropriately implemented and integrated with the human operator, the vessel probably would not have grounded. Because of the Safety Board’s concern that automation on other vessels has not been appropriately implemented and integrated with the human operator, the Board believes that the Coast Guard should propose to the IMO that it apply existing human-factors engineering standards in the design of integrated bridge systems on vessels."

Operational design points for the collision ASH / DUTCH AQUAMARINE (from MAIB report)

Navigators appear to prefer to return the vessel to the original planned track rather than parallel it until the next way point as was common practice before the advent of GPS. Where a track control system is used, it is, possibly, even more likely that the navigator will want to return the vessel to the planned track quickly after any deviation. It should be noted in this context that Vopak’s fleet orders to navigating officers instruct them to ensure that “…..the GPS gives the autopilot a course to return the ship as quickly as possible to the track to be sailed between the set waypoints.” A graphic example of this occurred when Dutch Aquamarine reverted to her original track after passing Lapad at about 1505. Dutch Aquamarine reverted to her track despite the fact that this meant that she passed Hornestrand at 1512 at a distance of only about 0.15 mile (see Plan 2). Most GPS and track control systems have functions that enable a new course direct to the next waypoint to be selected and steered. It is apparent that many navigators might not be adept at using all the functions of these instruments. On Dutch Aquamarine, the second officer had received no formal training in the operation of the IBS, including the track control system.

Collectively, these features of modern marine navigation effectively keep vessels bunched on the same or similar tracks.

...one possible explanation of why the second officer did not readily notice it might be that it was partially obscured by the heading marker on the radar screen. Indications from the CNIS recordings show that Ash would have been right ahead, or nearly right ahead, as Dutch Aquamarine approached. Every radar has the facility to temporarily switch off the heading marker on the screen so that operators can detect any targets that might be obscured and it is good practice to do this frequently during a watch.

Comment: There may be design solutions to preventing heading markers obscuring radar targets rather than requiring procedural fixes. The difficulty of parallel indexing is mentioned above in the problems section.

Cross Checking Positions MARS Report 200524

http://www.nautinst.org/search/mis/queryMARS.idq?srchString=200524

“The errors are obvious but I shall point out the main ones anyway:

1. Initial GPS alarm was cancelled when going into DR mode but no action taken, and the fact was not handed over or investigated.

2. Subsequent watches did not identify that the GPS or ECDIS were in DR mode despite ECDIS showing this fact. Personnel were not sufficiently familiar with the equipment.

3. The two independent GPS sets were not routinely compared.

4. Genuine and routine fixes were not taken from land. Where possible this should have been their primary fixing method only backed up by GPS! How the two fixes from land that were plotted were obtained is any ones guess, as they agreed with the DR track!

5. Radar tuning and adjustment of clutter would have made the OOW realise that this was in fact land he was detecting. At the very least it should have made him double check his position by other means and call the master.

Obviously it is a worry that the GPS can go into DR mode in such a way and mislead its user, perhaps with an inadequate alarm for this serious failure. Should it "re-alarm" systematically or not go into DR mode at all is perhaps something that should be discussed further by the governing bodies.

However, the real concern here is the lapse in basic navigational skills and the onboard management of the standard/company navigation procedures, with regards to the over reliance on a single GPS. Better use should be made of radar, echo sounder, common sense, experience and even the secondary GPS. All would have alerted the OOW to the danger not only immediately before scraping the rocky bottom but over the whole previous day.”

An alternative view is that these read very much like the classic problems of using people as monitors, and that the design may not have encouraged good practice; what can the INS PS do to reduce the risk of such (single person) error?

Comment by a Master who has done research on these topics:

"New Royal Majesty: Same events - GPS in DR, officer interpreting situation from the assumed world-view. But, unfortunately also same conclusions: people make mistakes. Lapse in nav skills and bad management. But it keeps happening..."

PRIDE OF PORTSMOUTH

MAIB Report on the investigation of the collision between Pride of Portsmouth and HMS St Albans Portsmouth Harbour 27 October 2002 Report No 20/2003 August 2003

“A fundamental prerequisite of good bridge management is that a mistake by any one person must not be allowed to go unnoticed and unchecked. The way that the bridge team was deployed, and the ergonomics of the instrumentation on Pride of Portsmouth, meant that monitoring of the master’s actions (by the chief officer and the helmsman) and those of the helmsman (by the chief officer and the master) was difficult and not effective.

To monitor the actions of the helmsman, it is necessary both to hear the master’s orders and the helmsman’s response, and to have sight of an RAI to ensure the correct action has been taken. The master did not have easy sight of an RAI from his conning position, and the chief officer could neither hear the orders easily nor easily see an RAI. Despite the fact that there was no misunderstanding between the master and the helmsman, this is considered to be a latent problem in the bridge management on Pride of Portsmouth becausethe inappropriate order was not noticed.

...

Additionally, both the master and the chief officer need a direct view of an RAI from their respective conning and monitoring positions. In the case of the chief officer on the bridge wing, this could be achieved by standing behind the existing control console, assuming an adequate view over the side can be obtained from that position. The master could not see the panoramic RAI situated directly above his conning position without moving away from the control console. This is an undesirable condition. The provision of an additional RAI in direct sight of someone operating the engine and thrusters controls from the central position would improve the safety and efficiency of bridge operations on Pride of Portsmouth.

...

In the crucial seconds before the collision the master became overloaded partly through giving verbal helm orders as well as having direct control of engine and thruster movements.”

GERDA

The Finish Accident Investigation Board Investigation report C 4/1998 M ‘ms GERDA, grounding outside port of Kotka, April 7, 1998’ appended to the incident analysis page.

This examined the Kelvin Hughes 5000T radar and its user interface.

The report considered the Kelvin Hughes user interface not functional for pilots because it differs from other radars. “The radars and steering equipment of a vessel the GERDA's type are analysed in accordance with the accident investigation of ms BALTIC MERCHANT (Investigation C 5/1998 M) based on piloting and the Rhein river traffic regulations. According to these regulations, the user interface of the Kelvin & Hughes radar is not acceptable. In the investigation of the BALTIC MERCHANT, the pilot clearly indicated that the user interface of the Kelvin & Hughes marine radar was slow to operate.”

Tasks examined included:

- adjusting radar image (e.g. gain)

- planning turns (e.g. EBL, VRM)

- use of parallel index

The Anschutz Nautopilot was also examined. Error in its operation by a single user may have contributed to the grounding.

Aratere

New Zealand Maritime Safety Authority Accident Investigation Report 04 3567 2005 Aratere near grounding - report appended to incident analysis page.

Norcontrol 2000 bridge equipment.

Conclusions included:

“The Kongsberg Maritime AS training/operational manual made no reference to the possible risk posed by the over reliant use of the automated systems.”

“There was a design flaw on the alarm system on the IBS, which alerted the bridge team only to sensor information failure that inputted into the IBS.”

The recommendations included:

“One possibility, for high priority alarms on the bridge, would be to introduce an alarm signal in which the navigation officer is alerted by a voice recorded alarm. This is not currently being considered in the revision of the above Code. The alarm could warn a navigation officer that the IBS has changed modes. This would eliminate any misinterpretation of what mode the IBS has switched into. Moreover, the officer would then be mindful of the mode change and take appropriate action as required.”

“It is recommended that Kongsberg Maritime AS, include in their future training manuals, a reminder to all owners, operators, and navigating officers, of the dangers of being over reliant on electronic navigational equipment, especially when operating in automatic-tracking mode.”

From the report.

The echo sounder was switched off.

It looks as though there was confusion over alarms and mode changes. “The alarm system on Aratere’s IBS alerted the navigation officer both visually and aurally to sensor information failures or XTE errors. The alarm system, however, fell short in that it did not warn a navigation officer that the IBS had changed its programmed steering mode, by way of an aural alarm. It simply indicated visually on the conning information display unit (Figure 6), that the IBS was in a different mode of operation. Nor did it warn the officer that the vessel was on a different track to that already programmed into the system. In short, there was nothing to warn a watch- keeper that human intervention was required for Aratere to alter course – in this incident, to prevent a grounding. In an attempt to ameliorate this shortfall, Interislander had installed an additional alarm system to alert the watch-keeping officer, by audible means, that the GPS link was suspect and therefore positions might be unreliable (this is one of the more common reasons for a steering mode change). However, in this incident, neither the Master nor the Chief Officer thought it was the GPS alarm that had sounded.”

Grounding of CONCERT EXPRESS – MARS Report No. 7022 offrep

With the speed vector superimposed over the heading marker, there is also the possibility that the pilot mistook the heading marker for the EBL, radar manufacturers may like to consider this when designing new sets and try to make the HM and EBL as visually distinctive as possible.

UK 2005

Recent near miss with a large tanker backing up to its berth, engine control was switched from the wheelhouse to the bridge wing conning position. Unbeknown to the bridge team the wing setting was for a speed substantially in excess of that on the wheelhouse control. The control did what it was told to do (silently and invisibly) and the ship accelerated accordingly toward shoal water and a very messy accident. Happily the individuals involved were sufficiently Situationally Aware and applied corrective action before the accident happened.

Aircraft automation

Flight Crew Reliance on Automation CAA PAPER 2004/10 Simon Wood ISBN 0 86039 998 2 Published December 2004 www.caa.co.uk

from the Conclusions section

1 Dependency on Automatics Leads Crews to Accept what the Aircraft is doing without Proper Monitoring

1.1 Summary

1.1.1 The availability of automation and automated decision aids encourages pilots to adopt a natural tendency to follow the choice of least cognitive effort. Training crews on automation bias or to verify correct automated functioning had no effect on automation-related omission errors, and neither did display prompts that reminded crews to verify correct functioning. However, there was evidence that pilots did perform better when the event was flight critical in nature.

2 Crews of Highly Automated Aircraft Lose Manual Flying Skills

2.1 Summary

There has been very little research published on the subject of the change in manual flying skill experienced by crews of highly automated aircraft. However, it is reported consistently that there is a discernible reduction in manual flying skills that is correlated both with the use of automation and whether the operation is long haul or short haul.

3 Inappropriate Response to Failures

3.1 Summary

3.1.1 Abnormal malfunctions have less well defined procedures compared with emergency situations and therefore crew revert to knowledge-based behaviour requiring more understanding of the system, plus time and effort to properly assess and resolve thesituation. This refocusing of tasks results in reduced levels of procedural accomplishment, communications and situational awareness, i.e. relatively minor failures can absorb both crew members in a way that is disproportionate to the significance of the problem. Published aircraft procedures specifically do not include elements of 'airmanship'.

3.1.2 System design is not always appropriate to keep pilots in the loop. (Research has shown that rather than design systems to work on thresholds or specific limits for control there should be a continuous flow of information to the pilot to indicate the difficulty or increasing effort needed to keep relevant parameters on target.)
to top