A Rationale for Future Bridge Navigation Displays
A new IMO Initiative
Prepared by Roy G Lee ( NavStans, UK )
______________________________________________________________________
In July 2000, the International
Maritime Organisation (IMO) at its NAV46 meeting agreed to a new
work item ‘to harmonize the presentation of navigation
information’ in such a way as ‘to avoid
confusion in the display of such information’ [1]. This
paper aims to give a brief review of some of the issues that need
to be addressed, especially in the area of distributed
information and clutter and to propose a means for moving forward
in a coordinated effort to solve the anticipated problems.
It is suggested that common integrated displays ( also known as
multi-function displays ) with data fusion - rather than
confusion - is perhaps the best way forward. Such common displays
should be duplicated ( or triplicated ) and should all be
‘Navigation and Hazard displays’ that give effective
‘situation awareness’ and ‘decision aid’
support to the mariner. They should be individually selectable
for the prevailing scenario and inclusive of all operational
needs, both for the displayed information and what is displayed
where. The selection of the presented information should be
filtered by the ‘need to know principle’ to enable the
user to reduce cognitive workload.
It is further strongly recommended that an Internationally agreed
‘common display surface’ to the user be defined in both
operation and symbology, irrespective of manufacturer. This is
considered to be the only safe way forward particularly with
predicted future manning levels. The maritime user should be
party to agreeing such an interface.
A regulatory impact assessment should be carried out so that all
stakeholders, including those involved in training, are better
informed of the benefits and risks.
After at least three years ‘on hold’ the IMO; spurred on by the recent agreement on the carriage requirement of the Universal Automatic Identification System ( U.AIS ) and to the relief of the International Electrotechnical Commission (IEC) working group on the Integrated Navigation System; has established a work item to discuss the harmonization of the presentation of navigational information. The U.AIS has its own ‘local display’ but this has not been specified under a best practice regime for maximizing the effectiveness of available information to the professional mariner. The IMO Performance Standard for the U.AIS specified an interface to potential graphic display surfaces but was silent on the actual requirements for such a display. With the U.AIS carriage requirement now agreed, the time is right for such a display surface to be specified and the IMO, not IEC, is the right forum to debate the issues involved. The real question is where is such new information ( effectively the reporting of the instruments and intentions of neighbourhood ships, Aids to Navigation (ATONs) and that relayed by VTS’s ) to be displayed – separately, co-located or combined?
The IEC Working Group 10, that has
been developing the IEC test specification for the IMO
Performance Standard for Integrated Navigation System (INS), has
known for some time that it would be wrong in principle for it to
specify ‘integrated display’ operations and symbology
without the further guidance of the stakeholders normally
involved in IMO negotiations. It also knew that these matters
were too important to be left to the market place. There was a
need for a top down operational specification or guidance for
Integrated Navigation displays [5], that is, a need for ‘A
Rationale for Future Bridge Navigation Displays’.
2. Distributed Information
– an Unsatisfactory Approach?
The information provided by navigational aids on the bridge of a typical ship is currently distributed over a number of ‘stand-alone aids’ - typically one or more radars; an electronic chart of some form ( i.e. an ECDIS or ECS ); a manoeuvreing or conning display as well as communications, machinery instrumentation and alarm panels etc. Distributing the information is not in itself bad design. But too often the task requires multiple instrument viewing and holistic comprehension by the mariner of multiple information sources. The difficulties are compounded by there being, currently, no consistency in the manner of operation or of symbology between the various display surfaces. Again if there are a sufficient number of trained personnel, acting in a coordinated manner, then distributed information can be an advantage – this may be the case in some military ships and in Vessel Traffic Systems (VTS) ashore. But manning is a costly resource and there is and will be, minimal availability in current and future ships. IEC has tried to provide, through the standards process, a methodology for harmonization of operation and control of navigational aids and the supporting display symbology, but the equipment implementations are many and varied. There is a burgeoning need for centralized control or an IMO specification or robust guidance.
The introduction of yet another navigational aid, namely the U.AIS, with its potential requirement of yet another stand-alone display, just adds to the difficulties of providing a satisfactory and safe system, without information overload. Thus it would seem that distributed information, with its possibility of yet more ‘heads down’ behaviour, is an unsatisfactory approach. Many accident reports also suggest that a decision aid with the joint hazard of collision and grounding on one display could be beneficial.
3. The Potential for Cluttered Displays
Perhaps some more information can be co-located or superimposed on to a current display. This has been done with mariner maps on the radar for a number of years, a ‘radar with selected parts of the System Electronic Navigation Chart (SENC)’ has recently been agreed and ‘radar as a layer in ECDIS’ is already an International Standard. These aids would have very cluttered displays were it not for the de-layering ( or the selected parts ) facility. The potential advantages of co-locating have been demonstrated in many available products such as the ‘added value’ provided by picture matching of the radar and chart coastlines. However, the displays can, for certain scenarios, become so cluttered that vital information can become ‘masked, obscured or degraded’ so that the ‘information content’ is unsafely reduced.
Indeed there are administrations, authorities and ‘experts’, who have declared that the stand-alone radar and electronic chart should be no more than that, i.e. that these should be pure radar or chart, unless it is part of the Integrated Navigation System (INS) and meets the relevant INS standard requirements.
The introduction, by co-location or superimposition, of the U.AIS display requirements ( even of selected parts of the U.AIS ) on to current display surfaces will certainly exacerbate the difficulties and it is unlikely that a satisfactory arrangement can be achieved that does not adversely affect the cognition of some displayed information.
At its Plenary meeting in 3Q99 the IEC opined, that because of the ‘seduction of safety’ that could be given by having U.AIS on ECDIS (the presented picture being incomplete), they would not support such a display combination. The same argument applied to a stand-alone U.AIS display. The only combination that IEC could put forward, for consideration, was the U.AIS /radar combination to aid the IMO functional requirement of collision avoidance – the presented picture should be complete for the task in hand. But use should also be made of ‘all available means’.
Fortunately any further exploitation of such co-location would violate the IMO Performance Standard for the INS where it says:-
Of course it may be that IMO at NAV47, in July 2001, may agree that this stipulation needs to be re-visited and re-negotiated under the new agenda item of ‘harmonization of the presentation of navigational information’. But ‘information overload’ be it by distributed or co-located information will need to be satisfactorily addressed and solved.
Information overload is dangerous in risky businesses such as maritime navigation – the military understand this problem with its potential for disaster only too well. One possible way forward is to combine data to provide data and information fusion and the reduction of the presented picture by the ‘need to know principle’ filter. Fusion and picture compilation, with rule based support, is well established in defence industries ( so there is nothing new or too costly in it ) [2] [7]. This is particularly relevant, certainly on larger ships, where there can be one or two radars pictures ( perhaps one on X- and the other on S-band ); a chart picture (perhaps with its own radar interlay card); a U.AIS transponder reported picture and a relayed VTS picture. All ostensibly giving the same picture, but all with their own distortions and non-observables - radar with rain and sea clutter outages; ECDIS because it is an historic picture; U.AIS because there will never be full transponder carriage; relayed VTS because the picture is compiled from a remote site. Thus knowledge of the characteristics and error budget of aids is needed in order to resolve image mismatch.
To illustrate the potential for display clutter consider the following simple scenario. That of 1, a visible ‘lighthouse’; 2, that is charted, that could at the same time also be conspicuous to radar and so has 3, a radar video return, it could also be 4, a radar tracked target and a reference target and carry 5, a U.AIS ATON transponder. There is of course just one lighthouse but the mariner has to ‘associate’, as the same object, five different presentations of the one real object. This association has to be done with the knowledge of the ‘normal mismatch’ that can obtain. Currently the four aids (2 ® 5) have different symbols that the user has also to remember from his training and infrequent re-fresh. How are these representations of a single object to be presented on a single screen without clutter? Although there have been some good attempts, they are not generally acceptable. Can they be safely associated ( or co-related ) and thus combined in such a way that a single symbol can be presented? Yes, the military and others have successfully done this using not just position but also all available spatial and temporal indicators. Choosing the best, rather than say Kalman filtering combining, using some rule-based support, can often perform such fusion at reasonable processing cost. This is particularly so now that GPS Selective Availability has been switched off.
It is noted further that the radar / radar plotting display has currently, for mainly historic reasons to do with the then available monochrome display technology, many different symbols to present various radar dependent existences ( i.e. acquisition, tracked, entering guard zone, CPA/TCPA warning, lost target, data requirements, ground reference etc. ) of a single target. These symbols perhaps need to be re-visited when combined on the same display screen as the U.AIS symbols ( which may need very similar entities ), but a policy of a single symbol for one real object recommends itself as a good rationale. Thus in general a single symbol should represent a single object with the characteristic of how it is made up and other entities should be un-intrusive both visually and to the fallible memory.
CIRM, at its 2Q2000 meeting, agreed to sponsor a literature search of what symbology on displays and controls and the definitions of related features and attributes that are extant. This was seen, by CIRM, as a necessary first step in harmonization of navigation information. Maximum use should be made of good extant symbology.
What concepts are available to guide the design of a good information filter?
‘Situation awareness’
has been used in the aviation industry for some time to describe
the top-level requirement of the instrument navigational aids to
the user. The definition most commonly used is that given by
Endsley [5] namely:-
In the maritime domain this defines spatial awareness and temporal awareness (of the near future) of the marine traffic and terrain with respect to navigational goals. Lack of situation awareness leads to poor decision making. It is recommended that situation awareness is a top-level requirement for an integrated display – it is already a requirement of the IMO INS. To aid the introduction of awareness concepts at the concept stage the designer should use a functional breakdown of tasks such as is in Ref [3] developed by ISSUS, BSH and DGON.
The preparation of an IMO input paper ‘to harmonize the presentation of navigation information’ will be aided by following the guidance of HEAP – Interim Guidelines for the Application of Human Element Analyzing Process ( HEAP ) to the IMO Rule-Making Process [8] [9]. This sets out a salutary checklist of questions of the technical, manning, training, management and work environment areas that need to be reviewed for an acceptable IMO instrument that incorporates the human element. It also emphasizes the necessary involvement of the maritime user in the process.
Persuasion of all stakeholders of the need for a new IMO instrument can only be done on the basis of clear and compelling need and have regard for the costs to the maritime industry and the burden of the legislative and administrative resources. A document setting out the regulatory impact assessment is thus also needed.
7. Tentative constraints and recommendations
The discussion on ‘presentation of navigation information’ suggests the following constraints:-
- that distributed information, with its possibility of yet more heads down behaviour is an unsatisfactory approach;
- the ‘added-value’ of co-locating information on a display can lead to cluttered results and that the ‘information content’ can be unsafely reduced, by being masked, obscured or degraded;
- information overload is dangerous;
- the only IMO mandated display surface for the AIS is the INS.
The discussion further proposes the following recommendations for consideration:-
These constraints and recommendations are proposed as ‘a means for moving forward in a coordinated effort to solve the problem’ of the harmonization of navigational information.
The author gratefully acknowledges the many discussion with colleagues at the Maritime and Coastguard Agency (MCA) UK and in many IEC/TC80 working group meetings, particularly those of the U.AIS and INS.
For completeness the following draft proposals for U.AIS symbology made during the margins of the May meeting of the U.AIS working group are repeated here, namely:-
- Consistency of symbology across all displays ( e.g. ECDIS display, radar ).
- Uniqueness – only one possible meaning.
- Non-ambiguity – ability to determine differences ( i.e. distinct ).
- Intuitively obvious ( e.g. pictogram ).
- Availability of information on appropriate displays.
- Each target shall have a basic symbol. Further attributes should be enhancements ( not changes ) to the basic symbol.
- The display of U.AIS information shall show a minimum size icon ( e.g. sleeping mode ) for all U.AIS targets.
- All U.AIS targets shall be displayed on initiation or after a pre-determined time-out ( e.g. default ).
The determination of what type of symbols are needed will be based on the functionality ( type and number of functions ) described in the IMO Operational Guidelines for the U.AIS.