Designing Situation Awareness

Addressing the Needs of Medical Emergency Response

Julia Kantorovitch

, Ilkka Niskanen

, Jarmo Kalaoja

and Toni Staykova

VTT-Technical Research Center, Espoo, Finland

CambridgeUniversity Hospitals, Cambridge, U.K.

Keywords: Situation Awareness, Design Principles, Decision Support, Medical Emergency Response.

Abstract: The effective support of Situation Awareness (SA) is the core of many applications. In this paper, we report

a progress on the research towards the complementing of the existing studies with new knowledge, on

engineering of SA in particular keeping in mind a complex multi-stakeholder context of existing and future

knowledge intensive intelligent environments. A medical emergency response use case is used as an

instantiation example to evaluate our engineering thoughts.

1 INTRODUCTION

The importance of Situation Awareness (SA) has

been firstly recognized for crews in the military and

aviation domains. The most prominent and complete

work in this direction is a Theory of Situation

Awareness studied and consolidated by Mica R.

Endsley (Endsley, 1995). In her work the theoretical

model of situation awareness and in particular its

role in human decision making is defined. From the

application point of view this research addresses

mostly the needs of various time-critical

environments, such as air traffic control, large

complex manufacturing systems, some medical

systems and tactical and strategic systems. These

systems are in a sense closed systems of single

provider supporting strict top-down design and

development approach and aim at addressing the

needs of a “single operator”.

On the other hand the increased availability and

robustness of sensors, the wide-spread use of the

internet, as well as intensified research in the area of

the content convergence and social media have led

to the definition and acceleration of various research

fields and phenomena that draw on the advances of

these technologies, Pervasive computing, Ambient

Intelligence and Internet of Things (IoT), as well as

cloud computing, which define a vision where in

the future distributed services and computing

devices, mobile or embedded in almost any type of

physical environment all cooperate seamlessly with

one another using information and intelligence to

improve user experience. The support for Situation

Awareness is equally important in this new context.

The value of SA can be found in the product

manufacturing domain, emergency management,

design, supply chain management, and equipment

remote maintenance, to mention a few applications.

The Situation Awareness requires applications to

support management of data, knowledge and related

services in an integrated and sustainable way.

Mastering of “simplicity” and “openness” will be

deterministic for the digital products, application and

services successful in the future. Openness in the

creation of products and services allowing various

independent networks of stakeholders to participate

in the services creation process will enable the

complementary bottom-up approach in smart

systems development. Simplicity is demanded by

users. Simplicity is related to usability and there is a

trend in industry to conflate these terms as much as

possible. Better usability will increase user

acceptance of technology, which is crucial for

products take-off in many contexts. The purpose of

the system may play an important role in defying of

respective elements that influence system acceptance

from the end-user as well as the developer and

business point of view.

In this research we aim at extending the existing

studies in the field looking on SA according this new

context. A medical emergency response context is

used as an example to instantiate and to evaluate the

respective design thoughts. Various stakeholders

Kantorovitch, J., Niskanen, I., Kalaoja, J. and Staykova, T.

Designing Situation Awareness - Addressing the Needs of Medical Emergency Response.

DOI: 10.5220/0006475504670472

In Proceedings of the 12th International Conference on Software Technologies (ICSOFT 2017), pages 467-472

ISBN: 978-989-758-262-2

467

dealing with Situation Awareness, such as end users

of the system (i.e. emergency responders), applica-

tion developers and business actors are examined.

By applying the user‐centred approach to elicit the

requirements for patient safety during emergency

medical response, one of our contributions to the

state of the art is to extract the “simplicity” hidden in

apparently complex emergency context. Our second

contribution lies in the resulted domain model and

respective knowledge oriented architecture, which

are defined to guide the development of desired

situation awareness.

In the following section, the related studies on

SA as well as the motivation for this research are

further discussed. The aim is to capture challenges

and needs still to be addressed in developing of SA

and therein to facilitate innovation and more

substantial system design. In Section 3, we discuss a

model based SA design for medical emergency

response context. Conclusions are summarised in

Section 4.

2 SITUATION AWARENESS

SA originated with aviation practitioners. After the

research has spread to other environments, such as

air traffic controllers, nuclear power plant operators,

anaesthesiologists and automobile drivers in

particular addressing cognitive tasks that operators

in such environments may face, thus extending the

theoretical model defined by Endley (Endsley, 1995

& Endsley, 2003).

Looking on the relevant advances in research in

the area of pervasive ubiquitous computing

environment and IoT, a related concept to SA is the

notion of Context Awareness (CA), which is defined

as ”any information that can be used to characterise

the situation of the entity” (Dey, 2001). The term

Context Awareness and Situation Awareness are

used interchangeably by some authors as they mean

the same, however there is an important difference

in their usage. The CA aims to enable better service

delivery through proactively adapting use and access

of information, physical resources and multi-

modality feature of human-computer interaction

process with respect to available context information

(Soylu, 2009; Schmidt, 2012). In contrast, SA is the

perception of the elements and “events” in the

environment related to the entity (i.e. user) or in

other words, it is simply about knowing and

understanding what is going on around. This

understanding may lead to some action taken by the

user.

While the methodology towards the developing

of context-aware applications for various intelligent

environments has appeared (Dey, 2001; Hong, 2009;

Perera, 2013), there has been a little focus put to

support the situation awareness in this new context

(Chen, 2012). The existing approaches aim mainly at

the addressing of development of domain specific

applications and improvements of existing

technology according to the theoretical models

introduced earlier (Endsley, 2003). For example, the

SA needs of an operator in road traffic management

domain are enhanced by developing a number of

ontology based applications (Baumgartner, 2010).

The machine learning algorithms towards the

recognition of situation are presented in

(Häussermann, 2010). A role of semantic

technologies in improving SA is discussed in

(Smart, 2007). In the domain of emergency

management, the research has been mainly focusing

on tackling of organizational aspects to achieve a

sufficient shared SA (e.g. Sapateiro, 2007;

Seppänen, 2013).

The existing approaches lacked the touch of

widely adopted software engineering practices

where the requirements to SA engineering are

mastered by examining various design-view points,

actors involved, and the domain specific aspects

such as standards and accepted work practices, as

visualised in Fig.1. Developing the effective and

sustainable Situation Awareness necessitates the

availability of respective Model, which is created to

build the description of the problem domain in

software engineering and to define the system

development process. Models are very much

associated with the domain they present.

Accordingly, in the following section we illustrate

our approach to the design of Situation Awareness

by researching and developing the domain specific

model to tackle the needs SA in a medical

emergency response context.

Figure 1: Design views in SA engineering.

ICSOFT 2017 - 12th International Conference on Software Technologies

468

3 EMERGENCY MEDICAL

RESPONSE

The information systems for emergency manage-

ment are based on information provided by various

actors, by diverse collections of sensors in the field

and information supplied by human volunteers. The

information may come in different forms such as

field reports, images and remote sensing informa-

tion. In order to achieve SA, various knowledge and

information models need to be aligned. It is widely

acknowledged that good SA leads also to good

decision making (Feng, et al. 2009). Moreover, as

various actors (and accordingly various heterogene-

ous information infrastructures) are involved as

information providers in the emergency management

context, there is a demand for the means to support

the interoperability among different information

sources towards their access and information reuse

and further acceptance of the system by business and

earlier adopters.

3.1 End-user View

The methods for the collection and analysis of

general emergency response user needs towards the

creation of domain model involved literature and

clinical practice reviews and face-to-face interviews

with stakeholders (e.g. the London Ambulance

Service, the Vienna Red Cross, the Sofia Military

Medical Academy). This led to the codification of

the principle five spaces, related actors and their

actions directly linked to medical emergency

response: (1) Initial Alert: the phase, where the

initial alert is being managed, usually a 112 call

center or Public Safety Answering Point (PSAP); (2)

Emergency Medical Service (EMS) on the Way: the

phase, in which an EMS team is dispatched to

emergency event’s location; (3) Field Management:

the event’s site where the people requiring urgent

medical help are located; (4) Transport: the phase, in

which an EMS team takes patients to a First

Receiver; (5) First Receiver: the phase, in which the

First Receiver, usually hospital, prepares for and

later takes over the care of the patient. The

emergency responders working in each of these five

phases/spaces have sets of patient-related tasks,

which are the same, irrespective of country or type

of incident. These are the tasks that form the basis

for the generic set of requirements for technology

and situation awareness and decision supports,

discussed next. The more the actions across the

spaces are interlinked by effective information

sharing technology and the more provisions for

mutual visibility, early situational awareness, and

decision support are provided, the more the phases

are enabled to run in parallel, therefore saving time

and becoming more effective in saving lives.

3.1.1 Situation Awareness Model

A Common Information Space (CIS) introduced to

maximise the quality of available information and its

outcome across five operational conceptual spaces

of emergency medical response, Decision support

points and Information sharing patterns constitute

the Situation Awareness model. The Situation

Awareness model is represented in form of the

sequence diagrams in Fig. 2. The purpose of these

sequence diagrams is: 1) to clarify the decisions that

different actors make during the course of an

incident and explain how the system supports deci-

sion making; 2) to illustrate the collaborative nature

of decision making and represent what kind of

critical information is required to support different

actors with their tasks; 3) to represent the important

information flows that potentially exist between

involved actors mediated by a knowledge mana-

gement system in form of notifications and alerts.

The emergency management process is started as

PSAP receives a call to the emergency telephone

number. Usually, the caller provides basic informa-

tion about the incident including the type and the

location of the incident and injured if any. PSAP

staff reports the received information to the CIS

system and determines also a priority dispatch code

for the event. The inserted information is transmitted

to the Decision support system that should be able to

analyse the data and to generate recommendations

for resources that should be invited to manage the

incident. The recommenda-tions are returned to the

CIS system, which in turn should notify PSAP staff

about the recommenda-tions. Subsequently, PSAP

staff may analyse the recommendations to make the

final decision about the resources that are invited

and dispatched. Next, PSAP staff communicates the

dispatch information to selected EMS staff members

and, additionally, informs hospitals through the CIS

system. After that, both EMS staff and the field

commander should compare the incident details (e.g.

the number of patients) against dispatched resources

and evaluate whether the required resource

estimations are accurate and justified. If necessary,

both of the aforementioned actors can decide to

dispatch additional resources for the incident. The

field commander also performs task allocation for

EMS staff members. The Decision Support system

may support this activity by creating

Designing Situation Awareness - Addressing the Needs of Medical Emergency Response

469

Figure 2: Sequence diagram of a collaborative decision making in an incident situation.

recommendations for task allocation utilizing, for

example, personal profiles that describe the

capabilities of involved EMS staff members. Actors

can also set critical tags to the system to indicate that

the incident involves potentially dangerous chemi-

cal, biological, radiological or nuclear materials. The

final decision making point presented in the Fig. 2

considers whether external experts should be invited

to the scene of an incident. Based on the incident

type, injuries of patients and/or the existence of

possible hazards and critical tags the field

commander may decide to insert information about

required external experts to the CIS.

The activities, decisions and communication

flows that are usually executed in the following

phases of an incident management process are not

illustrated in the sequence diagram due to the space

limit, however they are explained in the following.

Once EMS staff has examined the patients, patient

data (e.g. the results of triage) is reported to the CIS.

Next, the received information is analysed by the

Decision support system towards the generation of

recommendations for allocating patients to hospitals.

The recommendations are constructed by comparing

patients’ reported injuries against available hospital

data including provided specialities, location and the

number of available beds. Based on the received

recommendations EMS staff can decide the patient

allocation and is able to send allocation alerts to

hospitals and medical transportation. The allocation

alerts include the IDs of the patient and the hospital

receiving the patient. In the next phase, the CIS

system may communicate the location of hospitals

and current traffic information to the Decision

support system that should be able to process the

information and to generate route recommendations

for transportation vehicles. Once a patient arrives to

the hospital the hospital personnel downloads a

patient form from the CIS system thus

acknowledging the arrival.

Typically, the transportation personnel and

responders in the field also utilize specific cards that

offer guidance and thus facilitate the treatment of

different kinds of patients or allocation of required

resources.

The defined Situation Awareness model is used

as a basis to formulate the required knowledge

models for the Common Information Space and to

propose the overall architectural pattern that can be

used to achieve the desired SA functionality. The

ICSOFT 2017 - 12th International Conference on Software Technologies

470

proposed architecture and knowledge models, which

take into account the developer- and the business-

views are discussed next.

3.2 Developer and Business Views

Glossaries and vocabularies play a significant role in

emergency management due to the importance of

clear communication during disaster response. It is a

good practice to use standards if available to enable

information sharing interoperability. Several

standards addressing data modelling and data

exchange formats related to medical emergency

response have been designed so far. Based on the

conducted research (Kantorovitch et al. 2015),

OASIS EDXL based models (EDXL, 2015) appear

more promising to address the needs of medical

services in the context of emergency. To date, it is

the most complete and mature effort to facilitate

emergency information sharing and data exchange

across various actors - public, commercial and also

medical involved in the process of emergency

management. Consequently the EDXL-based

vocabularies have been selected as the central

knowledge models to utilize. Obviously there is no

possibility of universal agreement on any conceptual

scheme including EDXL, however it is argued that a

practical common ontology does not need to have

universal agreement, it only needs a large enough

user community to make it profitable for developers

to use it as a means to general interoperability, and

for third-party developer to develop utilities to make

it easier to use.

In addition, in order to support evolvability, the

system solutions have to take into account

requirements that arise from anticipated changes on

environment, technology and stakeholders’ needs.

Fortunately there is already accumulated knowledge

of well-known general software design principles

presented as architectural tactics and patterns that

can be used to guide design. Architectural tactic is a

characterization of architecture level decisions that

can be used to achieve a desired quality attribute

response. Evolvability is typically associated with

modifiability, which can be addressed by a tactic

localizing changes by increasing cohesion,

preventing ripple effects of changes by reducing

couplings, and deferring binding time to support

dynamic adaptability (Bachmann et al. 2007).

Studies have identified several architectural patterns

supporting evolvability, most important examples

being layering, Model-View-Controller (MVC)

pattern, and use of plug-ins (Bode and Riebisch

2010). Based on the best practices identified, the

architecture of the system is designed to adopt

layering and Model-View-Presentation (MVP)

architectural pattern that itself is a derivation of

MVC. Layers defined in MVP pattern are presented

with shades of blue in Fig. 3.

Model layer captures the information on problem

domain i.e. individuals of incidents and their

participants. The model stores dynamic data into

RDF store, which directly manages its logic and

consistency with axioms and rules. The Incident

ontology is the core ontology of Model. Model is

constructed using both domain specific and generic

ontologies shown as grey layers. The Incident

ontologies and EDXL concepts are structured in an

extensible way and constructed according to Linked

Data (LD) principles. Both, the developed Incident

ontologies and EDXL vocabularies are released as

an open source to GitHub software repository for

their further reuse (COncORDE, 2016).

Presenter layer typically retrieves data from the

model, and formats it to be useful in the views

(facilitated e.g. by REST API/JSON format). In the

emergency context the presenter layer contains

queries that report the overall situation and also may

enable alerts and notifications presented in Fig.2.

Presenter layer also supports queries and reasoning

based on generic ontologies for time, location and

organization or personal information. When other

external vocabularies are utilized by Model, new

presenters for them can be added. In addition,

Presenter Layers is designed to support other

functions and services of the system, such as

ontology-assisted information extraction, decision

support algorithms and overall management of

heterogeneous incident related content.

Figure 3: The knowledge oriented architecture.

Designing Situation Awareness - Addressing the Needs of Medical Emergency Response

471

Finally, a View can be any output representation

of information, such as a diagram or it may contain

multiple views, such as a map with several

information visualization layers.

Layers on the right shaded as green represent

stakeholder specific development, maintenance and

evolution related aspects of the system with

examples of supporting tools. For instance, ontology

evolution is supported with version control using

GitHub with wiki as a means for documenting the

rationales for changes in ontology. In order to ease

version control, ontology source is developed using

textual Turtle format.

The proposed framework utilizes two of the

main strengths of linked data technologies. First is

that the evolution of ontologies used by models can

be opened to collaborative work among developers.

Second is that the models themselves can be

extended and tailored for the specific needs of

systems and user views by choice of external

vocabularies and ontologies.

4 CONCLUSIONS

This paper has provided the detailed technical

description of a model based approach aiming at

achieving Situation Awareness and support for

decision making in a dynamic medical emergency

response context. The effective exploitation of

domain models, architectural tactics, linked open

data technology and domain specific vocabularies

aim at the interoperability, better acceptance and

evolution of the developed system. The use of Web

standards and a common data model makes it

possible to implement applications that operate over

the complete integrated data space. The focus of our

future work is put on further prototyping of the

proposed SA framework. The developed decision

support services are based on the mathematical

modelling of optimization problems for timely

allocation of resources and on semantically

supported domain knowledge modelling, as well as

on the machine-learning-based prediction of

emergency incident expected victims and

subsequently demand for resources.

ACKNOWLEDGEMENTS

This research is co-funded by EC in the context of

FP7 COncORDE project (607814).

REFERENCES

Chen, L., Rashidi, P. (2012). Situation, activity and goal

awareness in ubiquitous computing. International

Journal of Pervasive Computing and Communications,

Vol. 8 Iss: 3, pp.216 – 224

COncORDE incident ontologies and EDXL vocabularies

(2016). https://github.com/OntoRep/COncORDE

Baumgartner, et al (2010). Editorial: BeAware!-Situation

awareness, the ontology-driven way, Data &

Knowledge Engineering, v.69 n.11, pp.1181-1193

Bachmann, F and Bass, L. (2007) Nord, Modifiability

Tactics, SEI Techical Report

Bode, S. and Riebisch, M. (2010) Impact Evaluation for

Quality-Oriented Architectural Decisions Regarding

Evolvability, Proc. of ECSA conference, Copenhagen

Dey, A. K. (2001). Understanding and Using Context.

Personal Ubiquitous Computing, 5 (1), 4-7

Endsley, M. R. (1995). Toward a theory of situation

awareness in dynamic systems. Human Factors, 37,

32–64.

Endsley, M. R., Bolte, B., & Jones, D. G. (2003).

Designing for situation awareness: An approach to

human-centered design. London: Taylor & Francis

Feng, Y-H. (2009). Modelling situation awareness for

Context-aware Decision Support, Expert Systems with

Applications 36, 455–463

EDXL (2015). OASIS emergency management standards.

Häussermann, K., et al. (2010). Understanding and

Designing Situation-Aware Mobile and Ubiquitous

Computing Systems.Vol.14 No.1, pp.329-339

Hong, J., Suh, E., & Kim, S-J. (2009). Context-aware

systems: A literature review and classification. Expert

Systems with applications 36 (1), 8509-8522

Gartner, M. (1986). Situation awareness: Let’s get serious

about clue-bird, Draft.

Kantorovitch, J., Milis., et al. (2015). Knowledge

modelling framework - towards user-centered medical

services delivery in the emergency context. In Proc. of

ICT-DM conference, Rennes, Brittany, France

Sapateiro, C (2009). An Emergency Response Model

Toward Situational Awareness Improvement In Proc.

of ISCRAM – Gothenburg, Sweden

Schmidt, A., (2012). Context-Aware Computing Context-

Awareness, Context-Aware User Interfaces, and

Implicit Interaction. Online education, retrieved 2017

Seppänen, H., et al. (2013). Developing shared situational

awareness for emergency management, Safety Science

55 (2013) 1–9

Soylu, A. et al. (2009). Context and Adaptivity in

Pervasive Computing Environments: Links with

Software Engineering and Ontological Engineering.

Journal of Software, 992--1013

Smart, P. R. et al. (2007). Semantic Technologies and

Enhanced Situation Awareness. 1st Annual

Conference of the International Technology Alliance

(ACITA), Maryland, USA.

Perera, C., Zaslavsky, A., Christen, P., Geogakopoulos, D.

(2013). Context Aware Computing for the Internet of

Things: A Survey, pp. 1–41

ICSOFT 2017 - 12th International Conference on Software Technologies

472