Professional Documents
Culture Documents
DOI: 10.12716/1001.12.02.20
ABSTRACT: Situation awareness is the key element of the decision‐making process. In navigation safety
domain, monitoring, control, assessment of dangerous situations, support of operators of decision‐making
support system should be implemented in real time. In this paper we present the problem of situation
awareness applied to navigation safety control. The paper considers existing models of situation awareness and
ontology‐based approach for maritime situation awareness. We introduce the situation concept using infons.
Finally, we give an example how use these methods for design and creation of decision‐support system for
navigation safety control.
383
cause of incidents are not technical fault or damage, emphasize the notion of T.Gruber: “An ontology is an
but the human errors [1]. An inadequate assessment explicit specification of a conceptualization” [3].
of the situation is the most common reason of making
the wrong decision. Therefore, the continuous The ontology creation process for any system
analysis of current situations and assessment of their includes:
influence on the whole maritime situation is needed. definition of the subject domain and ontology size;
consideration of variants of recurrent‐use of the
Safe navigation implies collection, processing and existing ontology;
identification of relevance of information, obtained listing of important concepts in ontology;
from heterogeneous sources. The information sources definition of classes and classes’ hierarchy on the
for dangerous situation awareness can be: AIS, radar basis of “is‐a” and “kind‐of” relations;
data, information about presence of navigational definition of class properties which are called slots;
hazards, closed areas, ice and hydro meteorological definition of the facet for slots.
situations. It should be noted that the information
volume increase considerably in heavy vessel traffic The typical components of the ontology are the
areas, in narrow waters and ports. Thus, the critical following:
information could be missed and importance of the 1 Concept is the complete collection of different
available information could be incorrectly evaluated. individuals that share the most common features
Ultimately, it leads to the wrong assessment of the which can be more or less significant. While
dangerous situation. It is worth pointing out that the developing ontology of subject domain, the
capacities of the operator on assessment of the current definition of concepts’ structure of subject domain
situation are very limited, particularly in the case of and logical relations between them, their
the volume of acquired information. The operator can identification and formalization is being
not collect, analyze and interpret this inhomogeneous implemented. The final result of development of
information quickly and efficiently. For this reason it subject domain’s ontology is a hierarchy of classes
is necessary to develop new methods and containing the concepts of the subject domain and
technologies that will help operators in processes of relations between them.
collection, correct processing, situation awareness and Each concept is characterized by volume and
interpretation of results. content. The volume and content of concept is two
interrelated sides of concept. The volume is class
The major goal of this paper is the expansion of the of generalized objects in the concept, but content is
classical situation theory for maritime safety and collection of significant features by which the
description of the principles of organization of generalization and selection of these objects in this
collaborative decision making system for maritime concept is made.
safety. Also, we present a new method of situation 2 Relations integrate classes and describe them. The
awareness for maritime domain. This method allows most common type of relation used in ontology is
to describe different maritime situations such as categories of relationships. This type of relations
vessel collisions, ecological dangers, piracy, has several other names such as taxonomy
navigation under very difficult conditions (ice relation, “is‐a” relation, class‐subclass, hyponym‐
conditions, meteorological conditions, shallows) in hyperonym, subsumption relation, a‐kind‐of
terms of the infons theory. relation.
3 Axioms determine the conditions for correlation
between categories and relations. They express
obvious statements, which are connected concepts
2 ONTOLOGY FOR NAVIGATION SAFETY and relations. Axiom is a statement, which is
DOMAIN entered in ontology performed, and another
statement can be obtained from it. They allow to
Initial data for maritime situation awareness process show the information which can not be reflected in
is information about vessels, sea objects and ontology through creation of concept (class)
environment surrounding them. These data are hierarchy and different relations between concepts
compiled from different sensors (mechanical, natural) (classes). Axioms allow to take forward the
and provide information about properties and reasoning within ontology. In addition, based on
characteristics of monitored vessels and others sea axioms, we can form new rules which allow to
objects. The ontology of knowledge representation automatically add new contextual information into
about maritime objects needs to be developed for ontology. Axioms can constitute of limitations
realization of the formal approach of data collection imposed on some relations and allow to create
from sensors, which monitor environment. The new statements.
ontology includes information about objects and In addition to major components listed above the
relations between them, reflects any valid object ontology includes the instances that are concrete
changes and also measures their impact on the elements of some category.
current situation. In addition, the ontology must be
complete for realization of the real decision‐making The basic requirements for ontology of
system in navigation safety domain. collaborative decision‐making system for navigation
safety domain according to [4] are:
The notion “ontology” had first been proposed by simplicity: statements and relations need to be
R. Goclenius and used to refer to the branch of simple to use;
philosophy that studies the fundamental basis of flexibility and scalability: the addition of new
existence. Different notions of ontology in computer concepts and relations in ontology must be clear
science are given in [2] among which we can be and accessible;
384
universality: ontology should support the different In this paper situation awareness is, therefore,
kind of contextual information in navigation safety understood to mean continuous obtaining of
domain; information about the environment from
expressivity: ontology should support the heterogeneous sources in order to provide safety,
description of necessary number of attributes for integrating of this information into knowledge about
contextual information in navigation safety the current situation and analysis of their impact on
domain. safety, forecasting of future situations and assessment
their influence on safety. The situation awareness
The core of ontology for navigation safety domain process lies in understanding of surrounding
is abstract class Situation. Subclasses of this class are environment and what is going on around for
Goals, Objects and Relations. Entities of the subclass awareness of how information, situations and their
Objects can be physical or abstract and can have impact influence on the concrete goals and tasks in
characteristics (Attributes). They also can participate in the present and near future. Significant factors, that
relations. The class Attributes define specific object influences the on the correct and timely decision
characteristics such as location, speed, course, vessel’s making in case of dangerous situations, is the
name. The PhysicalObject subclass is a specific type of availability of complete, accurate and relevant
Objects subclass which is characterized by following information to decision makers. However, the
attributes: speed, draught and location. Relations decision making process in real systems for the event
subclass define the relation between sets of Objects. of maritime dangerous developments is complicated
For example, inRangeOf(X,Y) relation means that one by such specific factors as time limit for decision
instance of PhysicalObject X is within the range of making, accuracy and relevant information limit,
second instance of PhysicalObject Y. An important incorrect, unexpected and sudden‐onset situations
aspect of Attributes and Relations classes is that they which change in real time.
include two values which can be changed over time.
Each of these classes is connected with PropertyValues One of the first model for situation awareness was
class and is defined by two time dependent functions, proposed by M. Endsley [12] and includes three level
the first characterizes the current value and the other of the information processing:
characterizes the changing over time value. A new Level 1 (Perception of elements in the
PropertyValues is generated for Attributes or Relations environment) – perception of the current situation,
when Event comes on affecting Attributes or Relations. its properties and dynamic development of
Therefore, the value of Attributes or Relations can be elements, related to observed situation, in
defined on request to PropertyValues at any time environment.
(current, past or future). Event contains information Level 2 (Comprehension of the current situation) –
about environment obtained from different synthesis of the disconnected elements of the
heterogeneous sources (sensors) at some point in time situation received on the Level 1, comprehension
that influences specific Attributes and Relations. Event and processing of information, integration of
contains the specific entities, which signal changes of various heterogeneous data and determination of
current situation and thus are the means by which the its significance for particular situation.
situation representation evolves. Level 3 (Projection of the future status) – available
prognosis of future actions and future situation
development based on knowledge of situation
status and its elements’ behavior, for taking timely
3 SITUATION AWARENESS FOR NAVIGATION decisions on future actions.
SAFETY DOMAIN
Nevertheless, it should be noted that the problem
integration of heterogeneous information into a single
3.1 Situation awareness comprehensive picture of the environment at the
In the early 1980s of XX century, the situation theory semantic level of human comprehension and
was developed by J.Barwise and J.Perry [5–7] and projection remains open.
then was successfully extended by K.Devlin and
D.Pospelov [8, 9]. The situation awareness term has
many definitions and understandings. For the first 3.2 Situation awareness for safe navigation
time this definition had been introduced by
M.Endsley. According to the classical definition, Maritime situation is, therefore, understood to
situation awareness is “the perception of the elements represent a set of parameters which directly or
in the environment within a volume of time and indirectly, defined status of the monitored maritime
space, the comprehension of their meaning, and the object at the moment. Maritime situation awareness is
projection of their status in the near future” [10]. This an instrument for analysis of specific characteristics
basic definition was added by C.Dominguez at al. [11] and parameters of the monitored maritime object for
under which situation awareness includes the four the purpose the obtained information about its
components: current status and forecasting its status in the near
information about the environment; future [14]. In turn, the term maritime situation
integration of this information with internal management implies focused influence on navigation
knowledge of relevance for mental representation safety decision support system in order to improve
of this situation; the quality of decision making. This improvement is
application of this representation for future study in large part attributable to the changing of
of current situation in perception cycle; properties, characteristics and parameters of the
forecasting the future situations. system.
385
Initial data for maritime situation awareness navigation safety any activities can have a
process are information about vessels (vessel type, considerable effect on further development of the
draught, stability, floodability), location of areas with current situation.
solid ice cover, hydrometeorological conditions,
location of the closed areas for navigation, location of Any collaboration decision‐making system for
the environmentally vulnerable areas and others. navigation safety can be split into two subsystem:
knowledge management subsystem;
For situation awareness in navigation safety navigation safety subsystem.
domain, it is necessary to have the following:
information about environment: location of the Knowledge management subsystem is intended to
neighboring vessels, relationship among vessel manage the whole knowledge domain while safety
tracking management system and other vessels, subsystem is implemented for specific monitored
hydrometeorological conditions, status of the object, in this case, for navigation safety. The main
marine environment, tide water, steams. tasks of the knowledge management subsystem are
information about technical and information tools the following:
of the vessel such as GPS, AIS, radar, gyro‐ data analysis of monitored object: it carries out
compass, fathometer, indicator. collection of primary data which are obtained from
information about spatial orientation: there should different heterogeneous sources (sensors, radars,
be real‐time information about dangerous etc.) and its transformation into ontological format;
maritime situations. event management: it generates useful information
time management: it is necessary to have lead time for monitoring situation on the object in order to
for decision making in case of the dangerous ensure safety. There are plan to provide the
maritime situation. analysis of retrospective information about
dangerous situations and current information
As can be seen from the above situation awareness about situation on the monitored object.
for navigation safety implies efficient and sufficient
assessment of the current situation as well as prompt Navigation safety subsystem allows to:
forecast of evolution of this situation in the context of implement maritime situation analysis and
environment. realization of the most appropriate crisis scenario
of current situation development (for example,
According to [13] the information about situation vessels’ collisions, maneuvering in cramped
can be formalized in the terms of infons. Infon is conditions);
defined as follows keep logs on the monitored object and give an
alarm signal if necessary;
i R, a1,..., an , provide retrospective information about
, dangerous situations on monitored object at the
users’ request.
where R is n‐place relation; a1 ,..., an are objects Assessment of the dangerous maritime situations
appropriate for R , that is objects appropriate to the lies in recognition of the following potentially
same types as given relation; is polarity of the dangerous situations:
infon. If 1 then objects stand in the relation R , rounding of closed zone for navigation;
if 0 — otherwise. If we apply operations of crossing of the border of traffic separation
conjunction, disjunction and situation‐bounded schemes;
quantification to infon, it will allow us to fuse several navigation hazards;
simple infons into a compound infon. Relationships shallow water;
among situations and infons is called supports collision with other vessel;
relationships. For given infon and situation s navigation in dangerous conditions (solid ice
statement s | can be written, which means that cover, storms, heavy swell, frog, etc.).
situation s supports infon .
From this variety of situations for each object sets
Distinctive feature of the navigation safety system of parameters are formed that describe the level of
is recognition of the situation types. System can danger for this object along predictable vessel’s
recognize the various types of objects, various types traffic. The whole information about navigation
of the relations, various types of activities, ect. The hazards, shallow water zones, closed zones for
basic types for situation awareness are the following navigation, borders of traffic separation schemes
[8]: the type of a temporal location, the type of a collaborative decision making system in navigation
spatial location, the type of an individual, the type of safety domain can be obtain from Electronic Chart
an n‐place relation, the type on an infon, the type of a Display.
type, the type of a parameter, the type of a polarity.
Currently, the existing Electronic Chart Display
For an object x and a type T can be written as: can be divided into three groups:
electronic chart display and information system
x :T (ECDIS);
electronic chart system (ECS);
that means that the object x is of type T. raster chart display system (RCDS).
It should be noted that decision for real problems Electronic chart display and information system is
of navigation safety is often made in complex and equivalent of the modern paper navigation charts
rapidly changing environment, thus, decision making within the requirement of the Regulation V/20 of
under conditions of uncertainty becomes paramount SOLAS Convention [15]. ECDIS provide
importance. For example, the solution task of the characteristics and parameters of chart objects such as
386
guidance, dangers, isobaths, the prohibited and cover (Y) and can not proceed with navigation.
limited areas for navigation and details about the Infon for this relation is:
conditions along vessel’s route for skippers at their
request. In addition to the traditional tasks of ECDIS IceDangerVessel |=<< isThreat,Ice,Vessel,1 >>
that include provisional and executive charting,
correction of the current location, new ECDIS tasks
4 isMoving(X): it points out a unary relation, where
are the assessment of navigation safety, correction of
the single parameter X denotes a vessel in
an electronic chart, organization of advance alarm etc.
movement. The infon for this relation is the
ECDIS displays accuracy cartographic data in real
following:
time in conjunction with current location of vessels
obtained from DGPS, GPS. System also processes and
reports information from other navigation sensors, for IceDangerVessel |=<< isMoving,Vessel ,1 >>
example, hyro‐compass, lag, echo sounder, radar.
Fig. 1 displays the major components of ECDIS. Another graphical show of infons for
IceDangerVessel situation is given on the Fig. 2.
The case study considered in this paper involves the 2 isFog(X,Y): it points out a binary relation where
following scenarios of dangerous maritime situations: the first parameter X is the current weather and
navigation along the Northern Sea Route in the second parameter Y is the vessel. Here the
complex ice situation; vessel navigates in fog‐bank conditions. The infon
navigation in complex hydrometeorological for this relation is:
conditions (heavy wind, frog);
oil spill as a result of two oil tankers’ collision. WeatherDangerVessel |=<< isFog,Weather,Vessel ,1 >>
The situation Navigation along the Northern Sea
Route in complex ice situation is called Fig. 3 shows different representation of infons for
IceDangerVessel. For this situation the following WeatherDangerVessel situation.
relations can be separated:
1 isNear(X, Y): it means the vessel is near an ice
cover which influences the future safety of
navigation with current values of speed, course
and direction. In this relation, the first parameter X
denotes an ice cover, and the second parameter Y
denotes a vessel. For this relation infon can be
written as:
Figure 3. Graphical show of infons for WeatherDangerVessel
situation.
IceDangerVessel |=<< isNear, Ice,Vessel ,1 >>
The situation Oil spill as a result of two oil tankers’
2 isClash(X, Y): it means collision of a vessel (Y) with collision is called OilSpillArea. In case of this situation
solid ice cover (X). Infon for this relation can be it is necessary to correct the planned route while
written as: taking into account assumed rounding the oil spill
zone. The relation involved is:
IceDangerVessel |=<< isClush, Ice,Vessel ,1 >> 1 isNear(X, Y): it means that the vessel X is near a
zone of the two oil tankers’ collision and it
observes the oil spill. The navigation safety with
3 isThreat(X, Y): the occurrence of a given event
current parameters (speed, course, direction) is
suggests that the vessel (X) is shackled by the ice
387
impossible. This relation points out binary Finally, it is appropriate to note that applying the
relation, where the first parameter X denotes the situation theory should be a priority in research
vessel and the second parameter denotes the zone where the initial data flows are unstructured, have
of the oil spill. The infon for this relation is: large volume and incorrect decision making leads to
fatal consequences. Also, this theory permits the
OilSpillArea |=<< isNear,Vessel,OilSpillArea,1 >> constant operative control for current situation and
forecasting of the future situations taking into account
the given information and knowledge.
For each infons mentioned above we can define
rules which allow to describe the set of all dangerous
navigation actions in collaborative decision‐making
system in navigation safety domain. For development ACKNOLEDGMENTS
of the rule base of collaborative decision‐making
system, it is necessary initially to define physical and
This research was supported by the Russian
abstract classes of ontology for navigation safety
Foundation for Basic Research (RFBR): grant № 18‐07‐
domain.
01203 and was implemented within budget support of
SRW SPIIRAS № 0073‐2014‐0005.
5 CONCLUSION
REFERENCES
Onrush of information technologies and computer
hardware, automation of different kinds of physical [1] International Maritime Organization, URL:
and mental human activity implies development of http://www.imo.org
new information tools and technologies to help [2] Staab S., Studer R. Handbook on ontologies. Springer
people. It should be note that situation theory for Dordrecht Heidelberg London New York, 2003.
navigation safety presented in this paper is far from [3] Gruber T.R. Translation Approach to Portable Ontology
simple in terms of both understanding and Specification. Knowledge Acquisition Journal, 1993,
realization. However, it is aimed at dealing with vol.5, pp.199‐220.
complex perceptual and cognitive problems. In any [4] Korpipää P. and Mäntyjärvi J. An ontology for mobile
sphere of human activity the effectiveness of decision device sensor‐based context awareness. In: Proceedings
making in case of dangerous situations is directly of CONTEXT, 2003, vol. 2680 of Lecture Notes in
Computer Science, 2003, pp.451–458.
related to analysis of environment, collection of the
[5] Barwise J. Scenes and other situations. In: Journal of
large volumes of data in near real‐time. The modern Philosophy, 78(7), 1981, pp. 369 – 397.
decision support systems can exercise the [6] Barwise J. The situation in logic. In: CSLI Lecture Notes,
computerized manipulation of large volume of data no. 17, Standford University, California, 1989.
both on environment and on security for monitoring [7] Barwise J., Perry J. Situations and attitudes. MIT Press,
region. Vigorously developing of telecommunication Cambridge, MA, 1989.
systems, data transmission technologies allows to [8] Delvin K. Situation theory and situation semantics.
transfer these volumes nearly all over the world. Handbook of the History of Logic 7. Elsevier North‐
Nowadays, nevertheless, the main problem remains Holland, Amsterdam, 2006, pp. 601 – 664.
providing precisely those data which are required [9] Pospelov D.A. Situacionnoe upravlenie: teoriya i
praktika. — M.: Nauka, 1986. — 288 s.
users for effective decision making in current
[10] Endsley M.R. Design and Evaluation for Situation
moment. The technology of maritime situation Awareness Enhancement // Proceedings of the Human
awareness allows to offer users all necessary Factors Society 32nd Annual Meeting, vol. 1, 1988. — P.
information for navigation safety. This technology is a 97 – 101.
precursor of development of complete collaborative [11] Dominguez C., Vidulich M., Vogel E., McMillan G.
decision making system in navigation safety domain. Situation Awareness: Papers and Annotated
The major goal of this system is navigation safety and bibliography. Armstrong Laboratory, Human System
exchange of data from different heterogeneous Center, 1994, pp. 17–28.
information sources between all participants in the [12] Endsley M.R. SAGAT: A Methodology for the
decision making process. It may be useful to utilize measurement of situation awareness. Hawthorne, CA:
the multi‐agent architecture for modeling of the Northrop Corp.
[13] Kokar M., Matheus C.J., Baclawski K. Ontology based
collaborative decision making system. Such situation awareness. In: Journal of Information Fusion,
architecture allows to make the system works on 10(1), 2009, pp. 83–89.
solution for problem in concrete subject domain and [14] Popovich V.V., Smirnova O.V., Popovich T.V. Maritime
to increase the effectiveness of situation awareness by Situation Monitoring for Safe Navigation on Northern
applying, for example, the theory of fuzzy sets for Sea Route. In: TransNav, The international Journal on
analysis of dangerous situations. Application of the Marine Navigation and Safety of Sea Transportation,
situation theory and fuzzy methods of decision vol. 10, no. 3, 2016, pp. 433‐440, doi:
making in design of decision support systems allow: 10.12716/1001.10.03.08
to improve the process of operative navigation [15] Guidance on SOLAS Chaper – Safety Navigation, URL:
safety; http://solasv.mcga.gov.uk
to monitor the most dangerous maritime
situations;
to share data between different participants of
decision making process and navigation safety;
to make decisions in navigation safety more
deliberate and coherent.
388