=Paper=
{{Paper
|id=Vol-3137/paper18
|storemode=property
|title=Integration of Decision-Making Stochastic Models of Air Navigation System Operators in Emergency Situations
|pdfUrl=https://ceur-ws.org/Vol-3137/paper18.pdf
|volume=Vol-3137
|authors=Tetiana Shmelova,Kseniia Lohachova,Maxim Yatsko
|dblpUrl=https://dblp.org/rec/conf/cmis/ShmelovaLY22
}}
==Integration of Decision-Making Stochastic Models of Air Navigation System Operators in Emergency Situations==
https://ceur-ws.org/Vol-3137/paper18.pdf
Integration of Decision‐Making Stochastic Models of Air
Navigation System Operators in Emergency Situations
Tetiana Shmelova 1, Kseniia Lohachova2 and Maxim Yatsko3
1
National Aviation University University, Liubomyra Huzara ave. 1, Kyiv, 03058, Ukraine
2
National Aviation University University, Liubomyra Huzara ave. 1, Kyiv, 03058, Ukraine
3
National Aviation University University, Liubomyra Huzara ave. 1, Kyiv, 03058, Ukraine
Abstract
The authors present a new objective-subjective approach to conflict management to ensure
proper collaboration of different aviation personnel using decision-making (DM) methods in
Certainty, Risk, and Uncertainty. For improvement of outcomes of the collective solutions the
collaborative decision-making (CDM) models are used, when is difficult to make the final right
decision based on many factors influencing DM of different aviation specialists (pilot, air
traffic controller, flight dispatcher) during the development of an emergency in flight. The
Integration of Uncertainty Models to the Collective Model of CDM has been presented. An
example is presented - the building of individual models and CDM models for the pilot-in-
command, flight dispatcher, and air traffic control officer in bad weather conditions (lightning
strikes and thunderstorm activity) during the approach of aircraft. The optimal landing
aerodrome was chosen for landing in bad weather conditions.
Keywords 1
Integration of Stochastic and Non-stochastic Models, Collaborative Decision Making,
Decision Making under Uncertainty, Emergency, objective-subjective approach, individual
and collaborative models, pilot, flight dispatch, air traffic controller, objective and subjective
factors
1. Introduction
In the process of aviation industry development, safe air transportation is the core objective of the
functioning of the aviation system. But despite the rapid and constant growth from the point of
developed technologies applied in the operational processes of air traffic control, flight planning,
modern airplanes manufacturing, and improvement of hours flown by pilots and flight performance, the
number of aviation accidents does not decrease. Detailed aviation accident statistics are presented in
the Annual safety reviews of the European Aviation Safety Agency (EASA) and the guide to European
statistics "Statistics Explained" [1; 2]. According to the published data from the EASA, the number of
aviation accidents during the last years has grown [2]. There were 789 fatalities in aviation accidents
involving European Union (EU) - registered aircraft over the period 2016-2020 [1]. Most of the air
accidents dealt with general aviation, near 80% (as known, general aviation consists of all civil aviation
aircraft other than commercial air transport) [1]. As per EASA annual safety review, the number of non-
fatal accidents in 2019 was higher than the average 10-year period before [2]. The number of
occurrences depended on the size and load of the aircraft involved in the accident, and the complexity
of emergency situations. Most of the occurrences in 2019 were related to difficult meteorological
conditions [2]. Besides, the possible sources for such statistics may have next reasons [3; 4]:
1. Insufficient pre-flight preparation
2. Inadequate training of aviation personnel
3. Gaps in developed procedures and manuals, situational unawareness
International Workshop on Computer Modeling and Intelligent Systems (CMIS), May 12, 2022, Zaporizhzhia, Ukraine
EMAIL: shmelova@ukr.net (T. Shmelova); kseniialogachova@gmail.com (K. Lohachova); maxim_yatsko@i.ua (M. Yatsko)
ORCID: 0000-0002-9737-6906 (T. Shmelova); 0000-0002-3802-5540 (K. Lohachova); 0000-0003-0375-7968 (M. Yatsko)
© 2020 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
4. Availability of single guidance for multiple users for non-conflict decision-making (DM),
especially in emergency situations
5. The influence of the socio-technical environment and non-professional factors
(psychophysiological, individual psychological, socio-psychological) on the person
6. Psycho-emotional states of operators.
Most of these reasons belong to human factor. Human error is considered the first cause of accidents
[2; 3; 4]. Indeed, as many safety scholars affirm, human error is only an epiphenomenon, that is a
secondary phenomenon that occurs alongside or in parallel to a primary event. The real cause of
accidents which induces an error are human performance and limitations, poor teamwork,
organizational pressure on the professionals in the team to obtain unreasonable performance, faulty
human-machine interaction, and conflict interaction of professionals in important DM too [3; 4]. It’s
clear, human factor should be included for progressive and sustainable development and safety
enhancement of the complex aviation system. Safety is the functioning of certain operational processes
in the aviation systems with the objective of controlling the safety risks of the outcomes of hazards
during operation. International Civil Aviation Organization (ICAO) developing proactive approaches
for the safety provision based on risk and human performance assessment [5]. The one from the modern
concepts is the information and intelligent support of collective solutions of different specialists in an
emergency [5]. There are a lot of professionals involved in the provision of safety during flight planning
and operations process. They are flight dispatchers, flight crews, air traffic controllers, maintenance
staff, ground handling personnel, etc. Each of them plays a major role at different stages, as the safe
flight starts not only from the aircraft departure. They strictly follow the manuals and legal documents
approved in the field of their professional activity [6; 7; 8; 9].
The main actions of aviation specialists in accordance with the stage aircraft operation are presented
in the Table 1.
Table 1
The main actions of aviation specialists according to the stage of aircraft operation
Aviation specialist Stage of flight Operations
Flight dispatch (FD) Planning Responsible for flight planning and organization, for
choice of the optimal flight route, alternate aerodromes,
and proper fuel amount calculation for definite flights,
regulated by international and national documents,
orders, and instructions for the normal and abnormal
operational environment of aircraft [6]
Pilot‐in‐command Flight Is holding the full right of DM before departure and
(PIC) taking all responsibility in flight, following existing
aircraft flight and operations manuals, quick reference
handbooks (QRH) in case of emergency and abnormal
conditions according to the type of aircraft [7]
Air traffic control Safety of Ensures required aircraft separation minima established
officer (ATCO) flights of in each sector of airspace and provides flight crews with
aircraft in the assistance in emergencies, according to the instructions
control sector defined and approved within particular air traffic control
of airspace sector, by national laws, letters of agreement between
neighboring countries, and handbooks in case of aircraft
emergencies [8; 9]
Sometimes specialists from other fields, such as emergency and ground services, and medical staff
may be involved in joint DM (Figure 1). For example, the digital health and telemedicine in emergencies
are applied, allowing to invitie qualified medical personnel for a consultation in case of incapacitation
of one of the pilots or one of the passengers [10; 11].
Figure 1: The team of human‐operators involved in Collaborative Decision‐Making during the flight
Mostly, the complexity, content, particularity of documents, regulations of the professional activity
of each aviation personnel are different, which does not allow to develop the general algorithm of
actions for all aviation staff in a specific situation, especially for difficult in-flight conditions, where
the uncertainty, lack of information and time for DM take place. That’s why arises the conflict between
actions and decisions of involved staff who making the decisions at the same time for one situation.
Therefore, to ensure proper collaboration and conflict resolution between different aviation personnel
for improvement of outcomes of group decisions, it is necessary to implement new approaches for
conflict management and integrate DM stochastic and non-stochastic models for Air Navigation System
(ANS) operators, especially in an emergency [12; 13].
In the concept of the Global Air Navigation Plan developed by ICAO [14], the proper collaboration
is possible through the provision of air traffic management system (ATM) members with an
environment that ensures enough storage of significant information and its proper usage among ATM
system. In this environment, all members closely interact with each other in making common decisions
and achieve so-called Collaborative Decision-Making (CDM). This concept foresees the improvement
of the whole performance of the ATM system in general taking into consideration the individual
performance of ATM system members. This approach allows choosing the direction of action with
respect to selected objectives, based on decisions made by each participant, and information exchange
influenced those decisions, applying main DM principles [15]. That is why ICAO promotes the global
implementation of performance management principles, gradually transferring the existing ATM
systems to performance-based global ANS. The performance-based approach (PBA) is a mean of
establishing the performance management process and is based on three principles of obtaining
effectiveness of solutions [16]:
A strong and competent focus on desired results of DM
Conscious and rightly of DM
Reliance on real facts and data for rational DM.
Hence, according to the ICAO requirements, the meeting of the day-to-day performance in
operations may be achieved through the mechanism of Flight and Flow Information for a Collaborative
Environment (FF-ICE) Concept [17]. Concept FF-ICE defines air navigation information requirements
for flight planning, flow management, air traffic management, and trajectory management, and to be a
basis of the performance-based ANS [17].
In the monograph, “Intelligent Automated System for Supporting the Collaborative Decision
Making by operators of the Air Navigation System during Flight Emergencies”, was researched the PIC
and ATCO CDM during flight emergencies using deterministic models in order to achieve the
maximum synchronization of operators’ technological procedures. Collaborative Decision Models are
created for only two operators such as PIC and ATCO [18]. The ability to reach a general consent on
the desired outcome to be achieved by all ANS operators (FDs, flight crews, ATCOs, maintenance staff,
ground handling personnel) in terms of performance results is the basic condition for the successful
application of the approach for conflict management, especially in emergency situations [13]. The
application of classical DM criteria under uncertainty makes it possible to take into account the factors
influencing the choice and find collective solutions for several participants in the process [19].
The purposes of the work are:
1. Integration of DM stochastic and non-stochastic models of ANS operators in emergency
2. To build the individual DM models for the PIC, FD, and ATCO in an emergency and to
determine a collective solution for all operators-participants of the process.
3. To consider an example of building CDM models for the PIC, FD, and ATCO in conditions of
uncertainty when it is required to choose an optimal alternate aerodrome in bad weather conditions
before the approach.
2. The Integration of Uncertainty Models to Collective Model of Collaborative
Decision‐Making
For the effectiveness of DM of human-operators (H-O)’s of ANS in emergency situations the
integration of Stochastic, and Non-stochastic models has been proposed [4; 12; 13]. There are the next
main steps of the Method of the Integration of Stochastic models (DM under Risk and Uncertainty) and
Non-stochastic models (DM in Certainty):
1. Analysis of the development of the emergency situation (ES) and synthesis of DM models in
accordance with conditions, factors, potential participants, and stages of the development of the event
(Figure 2).
2. Building deterministic DM models using Network Planning methods and determining output
results (Figure 2):
critical time Тcr; Тmid; Тmin; Тmax
critical paths of performance of actions of ANS operators in ES
ambiguous situations S1, S2, and synchronization of H-Os actions using stochastic models
alternate situations S and optimal CDM using stochastic models.
Figure 2: The deterministic DM models in ES for 3 human‐operators
3. Analysis of procedures of the main technology, conditions of development of ESs, information
about problem and conflict situations, and using the effective integration of DM models:
If there is statistical information about the conditions of the development of the ES (the
presence of probabilities of the development of the ES), then DM models under Risk conditions are
applied and a decision tree is built (Stochastic Uncertainty DM models)
In the case of absence of statistical information and probabilities, there are the factors which
influence on the development of the situation, then DM models under Uncertainty are applied and a
DM matrix is built (Non-Stochastic Uncertainty DM models)
4. Structural analysis of the ES and building of the Stochastic Uncertainty DM model (decision
tree) according to conditions the development situation. Definition of the number of DM stages and time
of stage ti; relevant output data: Ai; pi; ui; βi (alternatives at each stage Ai; probabilities pi and outcomes ui
for each alternative Ai; added risks βk influence on DM according to stages of the development situation,
increasing threats due DM of H-Os. (Figure 3). Optimal solutions according to conditions of
development of ESs. Risk R defined as:
𝑅 𝐴 𝑚𝑖𝑛 𝐴𝑖 𝑚𝑖𝑛 𝑡 𝑝𝑢 𝛽 (1)
where ...
Ai – alternative decisions, А = {А1, А2, … Аi, …Аm};
pj – probability of development situation, p = {p1, p2, … pi, … pm};
uij - expected outcomes of alternative actions, U = {u1, u2, … ui, … um};
βi – added risk (increasing threats), depend from stages of the development situation; B = {β1, β2, …
βi, ...βm};
ti - time of stage of the development situation, T = {t1, t2, … t, …tm}.
{P}
{A}
{P}
{A} {U}=f(A; p; β, t)
Ai
Ai
{P}
{A} {U}=f(A; p; β, t)
Ai
Figure 3: The Stochastic Uncertainty DM models in ES with outcomes {U}=f (A; p; β, t)
4. Factor analysis of the ES and building non-stochastic DM model (Table 2) in accordance with
factors that influence DM as DM matrix:
Alternative actions А = {А1, А2 … Аi … Аm}
Factors influence on DM according to situation Λ = {λ1, λ2 … λj… λn}, j 1, 𝑛
Outcomes uij depending on actions / alternatives Аi , influence of variable (uv) or static (us)
factors - λj
For variable factors (uv) is required to consider the development of the situation - time ti (from Non-
Stochastic DM model) and probability pj (from Stochastic Uncertainty DM model). For static factors
(us) is required to consider the normative documents [6 - 9], procedures connected with development
of situation, type of aircraft, runway technical characteristics, characteristics of operators [20]. DM
matrix 𝑈 𝐴; 𝛬 with H-O alternative actions {А}, factors {Λ}and outcomes {U} include (Table 2):
𝑈 𝐴; 𝛬 𝑈𝑠; 𝑈𝑣 𝛬 𝑡, 𝑝, 𝑁 ,
where
Us – set of outcomes, influenced by the static factors;
Uv – set of outcomes, influenced by the variable factors;
t - time of development of ES;
p – probabilities of development ES;
N – normative documents according to development of ES.
Table 2
Matrix of DM in uncertainty, with variable or static factors
Alternative actions in ES Variable factors influence DM Static factors influence DM in
in critical situation critical situation
A1 uv11, ,uv1n us1,… usm
A2 uv2,… uv2n us2,… usm
A3 uv3,… uv3n us3,… usm
5. The optimal solutions are found using classical DM criteria under uncertainty (criterion Wald
(maximin), criterion Laplace, criterion Hurwitz, criterion Savage) [19]. The choice of criterion depends
on the type of aircraft, type of flight, the conditions for the development of the situation, the nature of
the emergency. The short characteristic of criterion and flights:
Criterion of Wald (maximin) - if this flight is performed for the first time:
𝐴∗ 𝑚𝑎𝑥 𝑚𝑖𝑛𝑢 𝐴 ,𝐵 (2)
where
Ai – alternative solution from set {А};
Bj – factor from set of factors {Λ}.
Criterion of Laplace - if this flight is regular:
𝐴∗ 𝑚𝑎𝑥 ∑ 𝑢 𝐴 ,𝐵 (3)
Criterion of Hurwitz – different approach using optimism-pessimism coefficient α.
For charter flight:
𝐴∗ 𝑚𝑎𝑥 𝛼𝑚𝑎𝑥 𝑢 𝐴 ,𝐵 1 𝛼 𝑚𝑖𝑛𝑢 𝐴 ,𝐵 (4)
where
α - optimism-pessimism coefficient, 0 ≤ α ≥ 1, 0 – extreme of pessimism and 1 - extreme of
optimism.
Criterion of Savage – recalculation result after flight:
𝐴∗ 𝑚𝑖𝑛𝑚𝑎𝑥 𝑟 𝐴 , 𝐵 , (5)
where
rij – loss matrix for recalculations after DM:
𝑟 𝐴 ,𝐵 𝛥 𝑚𝑎𝑥 𝑢 𝐴 ,𝐵 𝑢 𝐴 ,𝐵 (6)
6. Integration of the deterministic, stochastic uncertainty (DM in Risk) and non-stochastic
uncertainty (DM in uncertainty) models and determining optimal solution.
The flight pattern of an aircraft, the occurrence of an emergency (for example, when approaching
the destination airfield, weather conditions have changed, bad weather conditions (BWC) is shown in
Figure 4 with the following output data:
Flight – performed for the first time
Aircraft – Boeing -737
Aerodromes – take-off (A1), landing (A2) and alternate aerodromes (A2; A3; A4)
Emergency – BWC (lightning strike and thunderstorm activity)
The place of the event in-flight – emergency before approach
Figure 4: The scheme of aircraft flight route with possible alternative aerodromes
The decision for the selection of the alternate aerodrome in an emergency was implemented with
the following output data (Table 3):
The decision-making matrix
Alternative actions - {А} = {А1, А2, … Аi, … Аm} of DM H-O in ES and aerodromes of take-off
(A1), landing (A2) and alternate aerodromes (A2; A3; A4)
States of nature or factors {Λ}= {λ1, λ2, … λj, … λn} – variable or conditionally static factors {Λ}
and outcomes Us(Λ) influence on DM in ES. For example, DM in BWC (lightning strike and
thunderstorm activity); “variable factor” - are meteorological conditions on aerodromes; “static
factor” - available approach systems and aerodrome, crew, aircraft capabilities
Outcomes of DM matrix {U} = u11, u12, …, uij, …, unm.
Conditions of DM under uncertainty if this flight is performed for the first time - Criterion Wald
(maximin)
Table 3
Matrix of DM in uncertainty for solution optimal aerodrome in BWC
Alternative Meteorological Distance Fuel Aerodrome Crew Aircraft
aerodromes situation reserve capability capability capabilities
λ1 λ2 λ3 λ4 λ5 λ6
Take‐off aerodrome u11 u12 u13 u14 u15 u16
A1
Landing aerodrome u21 u22 u23 u24 u25 u26
A2
Alternate u31 u32 u33 u34 u35 u36
aerodrome A3
Alternate u41 u42 u43 u44 u45 u46
aerodrome A4
Alternate u51 u52 u53 u54 u55 u56
aerodrome A5
The expected outcomes of the decision matrix are formed based on the influence of factors on DM,
and the requirements of regulatory documents [6 – 9; 14 - 17], according to the Aeronautical
Information Publication (AIP) too. It is proposed to evaluate the outcomes of the decision matrix using
the Expert Judgment Method (EJM) [4; 19]. Each participant in the event - an aviation specialist fills
in the decision matrix based on personal opinion. All matrices have the same factors that influence DM.
2.1. The Algorithm of the Collaborative Decision‐Making in conflict /
emergency situation
The Algorithm of the CDM in conflict/emergency was obtained using the methods of DM under
uncertainty for effective collective solutions of different aviation specialists in an emergency:
1. Calculation of route direction
2. Building of DM matrix with:
Alternative solutions {А}
Factors, influencing on DM {Λ}
Outcomes of choosing of alternative solutions caused by factors influencing on DM {U}
3. Alternative solutions {А} - the list of alternate aerodromes (AA):
𝐴 𝐴𝐷𝑒𝑠𝑡 𝑈 𝐴𝐷𝑒𝑝 𝑈 𝐴𝐴 𝐴 ,𝐴 ,…𝐴 ,…,𝐴 ,
where
alternate aerodrome - an aerodrome of departure (ADep - A1) and it’s characteristics;
alternate aerodrome - an aerodrome of destination (ADest – A2) and it’s characteristics;
other alternate aerodromes and it’s characteristics according to the calculated route – A3; A4; A5, .
4. Factors {Λ} influencing on DM for each operator (O1 - PIC, O2 - ATCO, O3 - FD, and Oi - other
aviation specialists). These factors may be original or identical and objective (Table 3). For
example, next factors:
{Λ} =λ1, λ2 …, λj, …, λm,
where
λ1 – fuel reserve on board;
λ2 – remoteness of alternate aerodrome;
λ3 – runway technical characteristics;
λ4 – meteorological conditions on alternate aerodromes;
λ5 – the approach lighting system;
λ6 – available approach system;
λ7 – available navigation aids;
λ8 – aircraft performance characteristics;
λ9 – the presence of radiocommunication;
λ10 –air traffic intensity, etc.
5. Outcomes {U} - formation of possible consequences {U} influencing on the selection of AA in
the case of emergency:
{U} = U11, U12, …, Uij, …, Unm,
where
{U} - set of outcomes of DM matrix Uij (i =1, … m; j =1, …, n).
The possible consequences Uij are defined by means of using the EJM according to data from the
regulatory documentation and opinions of Oi operators (PIC, FD, and ATCO and other aviation
specialists) [4; 6 – 9; 14 - 17].
5. Formation of the matrixes of solutions for each operator. Formation of the matrix 1 of solutions
for the first operator (O1 - PIC) (Table 4).
Table 4
The DM matrix of DM in Uncertainty for operator O1
The matrix 1 Factors influencing on DM for operator O1 ‐ PIC
{А} λ1 λ2 … λj … λn
Alternative А1 u11 u12 … u1j … u1n
actions in А2 u21 u22 … U2j … u2n
critical … … … … … … ….
situation Аi ui1 ui2 … uij … uin
… … … … … … …
Аm um1 um2 … umj … umn
Analogically, DM matrix for the second operator (O2 - ATCO); the third operator (O3 - FD) and
other operators who are involved in this situation (Figure 1).
6. To choose the methods of DM under uncertainty (equations (2) – (6)) considering the conditions
of DM under uncertainty (type of flight, emergency, conditions of development of situation).
7. Finding optimal solutions for each operator using the criteria of DM under uncertainty: Wald,
Laplace, Savage, Hurwitz (equations (2) – (6)):
А1* = Aj(O1) - solutions of PIC A(O1),
А2* = Aj(O2) - solutions ATCO,
А3* = Aj(O3) - solutions FD.
Analogically, DM for other aviation specialists.
8. Formation of the collective matrix of solutions (Table 5), where:
{А} – alternate aerodromes;
{λ} – optimal opinions of all operators (O1 - PIC, O2 - ATCO, O3 - FD, and Oi - other aviation
specialists) from individual matrixes.
{u} – outcomes - optimal decisions of operators in accordance with the selected criterion / flight
conditions from individual matrixes Aj(O1); Aj(O2); Aj(O3)
Table 5
The DM matrix of DM in Uncertainty for operators
The Results of optimal solutions by all operators
collective
matrix
{А} Aj(O1) Aj(O2) Aj(O3) Aj(Oj) … An(On)
* * *
Alternate А1 u 11 u 12 u 13 … u*1n
aerodromes А2 u*21 u*22 u*23 … u*2n
… … … … …. … ….
Аi u*i1 u*i2 u*i3 u*ij … u*in
… … … … …. … …
Аm u*m1 u*m2 u*m3 … u*mn
9. Finding of optimal solutions for all operators using the criteria of DM under uncertainty: Wald,
Laplace, Savage, Hurwitz. For example, for criteria of Wald (2):
𝐴∗ 𝑚𝑎𝑥 𝑚𝑖𝑛𝑢 𝑚𝑖𝑛 𝐴 , 𝐴 𝑂
where
Ai – alternative solution from set {А};
Aj (Oj) – factors {Λ} – opinions of operators from individual matrixes.
The factors in CDM matrix are objective. For each case, depending on the conditions of the situation
and priorities of DM, a specific criterion has chosen.
2.1.1. The illustrative example of the Collaborative Decision‐Making in
emergency situation
Lightning strikes may affect airline operations because of costly delays and serious disruptions to
flight schedules [20; 21]. The Aviation Safety Network safety database gives the following lightning
strike accident statistics [22]: 29 occurrences with the contributory cause of the lightning strike,
including 14 losses of control. A lot of jet airplane lightning strikes occur while in clouds, during the
climb and descent phases of flight than in any other flight phase. The reason is that lightning activity is
more prevalent between 5,000 to 15,000 feet altitude. The probability of a lightning strike decreases
significantly above 20,000 feet. That’s why airplanes flying short routes in areas with a high incidence
of lightning activity are likely to be struck more often than long-haul airplanes operating in more benign
lightning environments [22; 23; 24; 25].
A lightning strike may result in [25]:
communication failure
electrical problems
emergency descends
pilot incapacitation.
Expect events:
pilot may be blinded by a lightning strike
navigation system problems.
The situation considered in this work: bad weather conditions - lightning strike and thunderstorm
activity closely to the destination aerodrome.
Initial data:
1. Airplane: Boeing 737
2. Route (Figure 5): Kyiv (Boryspil) (A1) - Odessa (A2).
3. Actual and forecasted weather conditions at the time of arrival to Odessa - thunderstorm activity
in the vicinity of aerodrome, heavy shower rain, hail.
4. Alternate aerodromes [19]:
Chisinau (A4);
Kharkiv (A5);
Alternate aerodrome, which can be chosen along the route: Dnipro (A3).
5. Factors influencing on DM for each operator:
{F} - factors considered by operator O1 (PIC);
{L} - factors considered by operator O2 (ATCO);
{Λ} - factors considered by operator O3 (FD).
Considering the interaction of the PIC, FD, and ATCO, it is necessary to understand their roles and
responsibilities. The FD is responsible for the planning of the flight, the PIC is directly responsible for
the safe execution of this flight, ATCO provides ATC service, information, and assistance to the crews.
Therefore, when making a decision, they analyze a general group of factors, cause the main purpose is
the completion of the task (flight from point A to point B), including risk analysis, but from different
points of view. And the final decision in flight makes the PIC.
For rational CDM, each operator involved in DM has analyzed and considered the current situation.
There are 3 operators in the CDM process: PIC (O1), ATCO (O2), FD (O3) (Figure 1, Table 1). Each
operator composed a matrix of decisions, where alternative solutions are alternate aerodromes for the
route Kyiv - Odesa, and each operator considers the same factors in the current situation, but with
different priorities.
The common factors for each operator (fj, lj, λj) taken into consideration when making a decision
when approaching the destination aerodrome:
f1, l1, λ1 – fuel reserve on board of the aircraft (always controlled);
f2, l2, λ2 – meteorological situation (of departure, destination, alternate aerodromes, en-route);
f3, l3, λ3 – aircraft capabilities (available equipment on board, MEL peculiarities, existing
operating limitations);
f4, l4, λ4 – aerodrome capabilities (available approach systems, technical characteristics of the
runways and taxiways, lightning system, service hours restrictions, aerodrome category, firefighting
and search and rescue category, emergency service);
f5, l5, λ5 – crew capability (crew operating minima, crew duty time);
f6, l6, λ6 – location of obstacles in approach, missed approach and departure sectors;
f7, l7, λ7 – air situation (intensity of air traffic control sector, radio frequency overload);
f8, l8, λ8 – commercial point (airport charges, distance from destination aerodrome, passenger
and cargo service facilities, the presence of contracts with handlers, the presence of customs, border
and migration control service, etc.).
Figure 5: The route of flight Kyiv (Boryspil) ‐ Odessa (UKBB ‐ Кyiv (Boryspil); UKOO – Odesa; LUKK –
Chisinau; UKHH ‐ Kharkiv (Osnova); UKDD‐ Dnipro)
These factors are objective. Composed the operators' DM matrixes when approaching to the
destination aerodrome in BWC (Tables 6-9). Expected outcomes considered by the PIC (operator O1)
represented in Table 6. The optimal solutions were determined according to the criteria Wald, Laplace,
Horvitz, Savage (2) - (5). The solution according to the criterion of Savage was obtained from the loss
matrix (6).
Table 6
The DM matrix of DM in Uncertainty for operator O1 (PIC)
Factors {F}, influence DM for operator Solutions
The matrix 1
O1 ‐ PIC
f1 f2 f3 f4 f5 f6 f7 f8 W L H, S
Alternative decisions {A}
α=0,5
Departure Kyiv (A1) 4 10 10 10 8 10 9 9 4 8,8 7 6
Destination Odessa
10 1 6 7 1 10 2 7 1 5,5 5,5 9
(A2)
Dnipro (A3) 5 9 6 7 7 9 6 7 5 7,0 7 4
Chisinau
Alternate 6 7 8 9 8 9 7 9 6 7,9 7,5 3
(A4)
aerodromes
Kharkiv
4 10 7 7 6 5 6 7 4 6,5 7 6
(A5)
The optimal landing aerodrome during the approach on the route "Kyiv - Odessa" in accordance
with the PIC's decision is as follows:
by Wald criterion - Chisinau (A4)
by Laplace criterion - Kyiv (A1)
by the Hurwitz criterion - Chisinau (A4)
according to the Savage criterion - Chisinau (A4)
Expected outcomes considered by the ATCO (operator O2) represented in Table 7.
Table 7
The DM matrix in Uncertainty for operator O2 (ATCO)
Factors {L}, influencing on decision‐ Solutions
The matrix 2
making of ATCO (O2)
Set of alternative l1 l2 l3 l4 l5 l6 l7 l8 W L H, S
decisions {A} α=0,5
Departure Kyiv (A1) 3 10 10 10 9 10 5 7 3 8,0 6,5 7
Destination Odessa
10 1 10 10 1 10 8 10 1 7,5 5,5 9
(A2)
Dnipro
4 8 10 10 6 8 6 5 4 7,1 7 6
(A3)
Alternate Chisinau
8 5 10 10 7 7 7 9 5 7,9 7,5 5
aerodromes (A4)
Kharkiv
8 4 10 10 7 7 4 8 4 7,3 7 6
(A5)
The optimal landing aerodrome during the approach on the route "Kyiv - Odessa" in accordance
with the ATCO's decision as follows:
by Wald criterion - Chisinau (A4)
by Laplace criterion - Kyiv (A1)
by the Hurwitz criterion - Chisinau (A4)
according to the Savage criterion - Chisinau (A4)
Evaluation of optimal alternate aerodrome for landing in difficult meteorological conditions by FD
at the stage of flight planning. Matrix of possible outcomes of DM by FD during choosing of alternate
aerodromes at the stage of flight planning represented in Table 8.
Table 8
The DM matrix for operator O3 (FD)
Factors {Λ}, influencing on decision‐making Solutions
The matrix 3
of FD (O3)
Alternative decisions λ1 λ2 λ3 λ4 λ5 λ6 λ7 λ8 W L H, S
{A} α=0,5
Departure Kyiv (A1) 3 10 10 10 7 10 7 5 3 7,8 6,5 7
Destination Odessa
10 1 7 7 1 10 3 10 1 6,1 5,5 9
(A2)
Dnipro
3 8 6 5 5 8 5 8 3 6,0 5,5 5
(A3)
Alternate Chisinau
8 9 8 7 5 5 6 9 5 7,1 7 4
aerodromes (A4)
Kharkiv
6 4 9 8 6 8 4 10 4 6,9 7 6
(A5)
The optimal landing aerodrome during the approach on the route "Kyiv - Odessa" in accordance
with the FD 's decision is as follows:
by Wald criterion - Chisinau (A4)
by Laplace criterion - Kyiv (A1)
by the Hurwitz criterion - Chisinau (A4) and Kharkiv (A5)
according to the Savage criterion - Dnipro (A3)
To determine the consistency of operators, collective matrices were constructed, in which the factors
in the decision matrixes for the operators (PIC (O1), ATCO (O2), FD (O3)) are identical, the solutions
of the operators and taken from matrices, presented in Tables 6; 7 and 8. In the CDM matrixes used
subjective factors – opinions of operators. The optimal CDM if this flight is performed for the first time
is presented in the Table 9 (Wald criterion). In this case, the optimal landing aerodrome, determined by
objective (fuel reserve on board of the aircraft; meteorological situation; crew, aircraft and aerodrome
capabilities; location of obstacles in approach; air situation and commercial point) and subjective factors
(PIC, FD, and ATCO) is alternate aerodrome Chisinau (A4).
Table 9
The CDM matrix for all operators (criteria Wald)
PIC ATCO FD CDM
Alternate aerodromes O1 O2 O3 Criterion Wald
Kyiv (A1) 4 1 5 1
Odessa (A2) 1 1 1 1
Dnipro (A3) 5 4 5 4
Chisinau (A4) 6 5 5 5
Kharkiv (A5) 4 4 4 4
The optimal CDM if this flight is regular (Criterion of Laplace) presented in the Table 10 – is Kyiv
(A1).
Table 10
The CDM matrix for all operators (criteria Laplace)
PIC ATCO FD CDM
Alternate aerodromes O1 O2 O3 Laplace
Kyiv (A1) 8,8 7,8 8,0 8,2
Odessa (A2) 5,5 6,1 6,1 5,9
Dnipro (A3) 7,0 6,0 6,0 6,3
Chisinau (A4) 7,9 7,1 6,8 7,3
Kharkiv (A5) 6,5 6,9 6,9 6,8
The optimal CDM in different approaches using optimism-pessimism coefficient α=0,5 (criterion of
Hurwitz) presented in Table 11 – is Dnipro (A3).
Table 11
The CDM matrix for all operators (criteria Hurwitz)
PIC ATCO FD CDM
Alternate aerodromes O1 O2 O3 Hurwitz
Kyiv (A1) 5,5 5,5 5,5 5,5
Odessa (A2) 7 7 5,5 6,3
Dnipro (A3) 7,5 7,5 7 7,3
Chisinau (A4) 7 7 7 7,0
Kharkiv (A5) 5,5 5,5 5,5 5,5
Collective decisions were analyzed with varying degrees of optimism-pessimism coefficient α=0,5.
The consistency of decisions increases with an increase in the coefficient of optimism, with a decrease
in the coefficient in the direction of pessimism, the mismatch increases.
The consistency of decisions determined using the Savage criterion (the recalculation after a flight),
were determined for cost initial matrix (Table 12):
𝐴∗ 𝑚𝑖𝑛𝑚𝑎𝑥 𝑟 𝐴 , 𝐵 𝑚𝑖𝑛 3; 6; 1; 0; 3 0 𝐴 ,
where
𝑟 𝐴 ,𝐵 𝛥 𝑢 𝐴 ,𝐵 𝑚𝑖𝑛𝑢 𝐴 ,𝐵 3; 6; 1; 0; 3
Table 12
The CDM matrix for all operators (criteria Savage) – recalculation)
PIC ATCO FD CDM
Alternate aerodromes O1 O2 O3 Savage
Kyiv (A1) 6 7 7 3
Odessa (A2) 9 9 9 6
Dnipro (A3) 4 6 5 1
Chisinau (A4) 3 5 4 0
Kharkiv (A5) 6 6 6 3
The loss matrix presents on the Table 13. The loss matrix shows risks if operators do not choose the
optimal collective solution. The maximum risks are selected, which then are minimized.
Table 13
The loss CDM matrix for all operators (criteria Savage))
PIC / loss ATCO/ loss FD / loss Max loss
Alternate aerodromes O1 O2 O3 Savage
Kyiv (A1) 3 2 3 3
Odessa (A2) 6 4 5 6
Dnipro (A3) 1 1 1 1
Chisinau (A4) 0 0 0 0
Kharkiv (A5) 3 1 2 3
The optimal landing aerodrome, determined by objective and subjective factors is alternate
aerodrome Chisinau (A4) as in Wald criterion. The calculations showed a balance between safety and
cost of the flight using criteria Wald (maximum safety) and criteria Savage (minimal cost and loss).
3. Acknowledgements
Thanks to participants of the scientific research - professional aviation specialists such as pilots and
dispatches from airline.
To demonstrate the new method of Integration of Decision-Making Stochastic Models and CDM
used the individual science works of aviation students in education (courses “Informatic of DM in ANS”
in National Aviation University, Kyiv). Students of different qualifications (PIC, FD, and ATCO,
operators of Unmanned Air Vehicles, engineers, and technical personnel) study theoretical courses and
then use their knowledge for scientific research. In the chapter "Collaborative Decision Making in
Emergencies by the Integration of Deterministic, Stochastic, and Non-Stochastic Models" [13] was
presented example of the emergency situation "Lightning strike" from a master's diploma of Zhanna
Maksymchuk's “Dynamic models of air traffic controller decision-making in difficult meteorological
conditions "Lightning strike" [25]. The diploma considers an example of choosing the optimal landing
aerodrome for the ATCO. The authors proposed to find joint solutions for the PIC, FD, and ATCO in
an emergency situation.
4. Conclusion
Collaborative Decision-Making is a process of presenting individual and collaborative information
and decisions made by various interacting participants, such as PIC, FD, and ATCO in professional
solutions, providing synchronization of decisions taken by participants, the exchange of information
between them, the effective balancing between safety and cost in collective solutions. It is important to
ensure the possibility of making a joint, integrated solution by aviation specialists at an acceptable level
of efficiency with minimal risk and maximum safety [5]. This is achieved by the completeness and
accuracy of the available information, and by the well-coordinated interaction between specialists, their
clear and correct understanding of job duties, and their roles in the process of completing a common
task. A Collaborative Decision-Making process in uncertain situations should be provided by using DM
stochastic models (stochastic uncertainty and non-stochastic uncertainty models) to improve and
simplify the deterministic model.
he objective-subjective approach is effective for determining the optimal solution in important and
difficult situations and in conflict interactions between operators. After analysis of the situation firstly
the performance of the synthesis (aggregation) of individual models (with objective factors), the next
step is to determine collective solutions (with subjective factors). The example of choosing an
aerodrome in case of an emergency (for example, in difficult meteorological conditions) using the
methods of DM under uncertainty was presented. The calculations showed a balance between safety
and cost using criteria Wald (maximum safety) and criteria Savage (minimal cost and loss).
The direction of further research is to develop the DM models for all CDM participants within the
Airport CDM (A-CDM) concept may aggregate the solutions of partners (airport operators, manned
aircraft operators, unmanned aircraft operators, ground handling agents, and air traffic services) in joint
work within non-segregated airspace [26]. For more accurate and timely information provision all
partners at the airport are required to create a single operational picture of air traffic [15; 17]. The
research was carried out during the simulator training of air traffic controllers [27], it is planned to build
the same models for other aviation specialists, operators of manned and unmanned aircraft, flights in
integrated airspace [28]. Specialists from other fields, for example, emergency and ground services, and
medical staff may be involved in joint DM. Priorities and effectiveness of connecting specialists in the
group for CDM will be in the next research. For effective CDM is proposed to include the participant
(Artificial Intelligence) in the group of participants in a joint solution and organize the research of a
hybrid optimal solution.
5. References
[1] Air safety statistics in European Union (EU), 2021. URL: https://ec.europa.eu/eurostat/statistics-
explained/index.php?title=Air_safety_statistics_in_the_EU
[2] EASA Annual Safety review 2020. European Union Aviation Safety Agency, 2020. URL:
https://www.easa.europa.eu/downloads/117065/en
[3] A. Chialastri, AF 447 as a Paradigmatic Accident: The Role of Automation on a Modern Airplane,
IGI Global, USA, Pennsylvania, 2021. doi:10.4018/978-1-5225-7709-6.ch006
[4] T. Shmelova, Yu. Sikirda, N. Rizun, Abdel-Badeeh M. Salem, Yu. N. Kovalyov (Eds.), Socio-
Technical Decision Support in Air Navigation Systems: Emerging Research and Opportunities,
IGI Global, USA, Pennsylvania, 2017. doi: 10.4018/978-1-5225-3108-1
[5] Safety Management Manual, 3rd ed. Doc. ICAO 9859-AN 474. ICAO, Canada, Montreal, 2013.
[6] Flight Operations Officer/Flight Dispatcher Training Manual, Doc ICAO 7192. ICAO, Canada,
Montreal, 1998.
[7] Procedure of making decision of departure and arrival of aircraft of Ukraine civil aviation
according instrument flight rules: State aviation authority of Ukraine Order No 295, 2005.
[8] Air Traffic Management. Doc. 4444-RAC/501. Fifteenth ed. ICAO, Canada, Montreal, 2007
[9] Guidelines for Controller Training in the Handling of Unusual/Emergency Situations (ASSIST). –
EUROCONTROL, Brussels, Belgium, 2003.
[10] Abdel-Badeeh Salem (Ed.), Innovative Smart Healthcare and Bio-Medical Systems: AI, Intelligent
Computing and Connected Technologies. 1st ed. CRC Press, 1st edition: December 28, 2020,
ISBN 9780367490614
[11] Emergency telemedicine and mission safety support, 2022. URL:
https://www.aircareinternational.com/inflight-emergency-telemedicine
[12] T. Shmelova, Yu. Sikirda, C. Scarponi, A. Chialastri, Deterministic and Stochastic Models of
Decision Making in Air Navigation Socio-Technical System, in Proceedings of the 14th
International Conference on ICT in Education, Research and Industrial Applications. Integration,
Harmonization and Knowledge Transfer, Kyiv, Ukraine 2018. URL: http://ceur-ws.org/Vol-
2104/paper_221.pdf
[13] T. Shmelova, Collaborative Decision Making in Emergencies by the Integration of Deterministic,
Stochastic, and Non-Stochastic Model, in: Jon W. Beard (Ed.), Information Technology
Applications for Crisis Response and Management, IGI Global, USA, Pennsylvania. 2021, doi:
10.4018/978-1-7998-7210-8.ch010
[14] Global Air Navigation Plan 2016-2030. Doc. ICAO 9750. Fifth ed., ICAO, Canada, Montreal,
2016.
[15] Manual on Collaborative Decision-Making (CDM). Doc. 9971. Second ed. ICAO, Canada,
Montreal, 2014.
[16] Manual on Global Performance of the Air Navigation System (PBA). Doc. 9883. First ed. ICAO,
Canada, Montreal, 2009.
[17] Manual on Flight and Flow Information for a Collaborative Environment (FF-ICE). Doc. 9965.
First ed., ICAO Canada, Montreal: ICAO. 2012.
[18] Yu. Sikirda, M. Kasatkin, D. Tkachenko, Intelligent Automated System for Supporting the
Collaborative Decision Making by Operators of the Air Navigation System during Flight
Emergencies, in: T. Shmelova, Yu. Sikirda, A. Sterenharz (Eds.), Handbook of Artificial
Intelligence Applications in the Aviation and Aerospace Industries, IGI Global, USA,
Pennsylvania. 2021, doi: 10.4018/978-1-7998-1415-3.ch003
[19] Hamdy A. Taha, Operations Research: An Introduction, Global Edition, 10th Ed., University of
Arkansas, 2019, ISBN-13: 9781292165554
[20] Rules of information provision of flight safety management system of aircraft of civil aviation of
Ukraine. Ministry of transport and communications Order No 295. 2009.
[21] Operation of Aircraft. ICAO Annex 6. ICAO, Canada, Montreal, 2018.
[22] The Aviation Safety Network, an exclusive service of Flight Safety Foundation, 2022. URL:
https://aviation-safety.net/
[23] G. Sweers, B. Birch, J. Gokcen, Lightning Strikes: Protection, Inspection, and Repair, Boeing
AERO Magazine. QTR. 2022. URL:
https://www.boeing.com/commercial/aeromagazine/articles/2012_q4/4/
[24] SkyBrary, 2022. URL: https://www.skybrary.aero/index.php/Lightning
[25] Zh. Maksymchuk, Dynamic models of air traffic controller decision-making in difficult
meteorological conditions "Lightning strike", Master’s thesis, National Aviation University. Kyiv,
Ukraine, 2020
[26] Airport collaborative decision-making, 2022. URL: https://www.eurocontrol.int/concept/airport-
collaborative-decision-making
[27] T. Shmelova, Yu. Sikirda, T. Jafarzade, Artificial Neural Network for Pre-Simulation Training of
Air Traffic Controller, in: Mehdi Khosrow-Pour (Ed.), Research Anthology on Artificial Neural
Network Applications (Volumes 3), IGI Global, USA, Pennsylvania, 2022. doi: 10.4018/978-1-
6684-2408-7.ch065
[28] T. Shmelova, V. Lazorenko, O. Burlaka, Unmanned Aerial Vehicles for Smart Cities: Estimations
of Urban Locality for Optimization Flight in: José António Tenedório (Ed.), Methods and
Applications of Geospatial Technology in Sustainable Urbanism (Chapter 15), IGI Global, USA,
Pennsylvania, 2021. doi: 10.4018/978-1-7998-2249-3.ch015