Flight Collision Avoidance System for Self-Separation Peter G Higgins Swinburne University Hawthorn 3122, Australia
[email protected] ABSTRACT
In a Free Flight Environment, pilots are delegated the authority to choose the flight paths, maintain spatial separation and consider environmental conditions with minimum intervention from Air Traffic Controllers. These factors constitute new tasks for pilots, which otherwise would be performed by Air Traffic Control. To maintain separation of aircraft, pilots are reliant on spatial awareness in controlling basic flight parameters such as speed, heading and altitude. Author Keywords
Free Flight; Ecological Interface Design; Cockpit Display of Traffic Information (CDTI); Situation Awareness; Human Machine Interaction; Self separation ACM Classification Keywords
H.1.2 User/Machine Systems: Human factors; H.5.2 User interfaces INTRODUCTION
A fundamental change to the management of air traffic is seen as necessary for alleviating the rapidly increasing demands on conventional Air Traffic Control. By adopting a Free Flight Environment [1, 2, 3], pilots supported by an Airborne Separation Assurance System (ASAS) would be delegated responsibility to create their own flight paths, with minimal interference from air traffic control [4, 5]. In managing their flights, pilots need instrumentation to help them detect and resolve conflicts. To perform selfseparation manoeuvres in a free flight environment, pilots need a supportive tool that clearly shows conflict geometries, and provides alternatives to overcome a loss of situation awareness of surrounding air traffic. An airborne collision avoidance display such as conflict display traffic information (CDTI) allows pilots to carry out “visual flight” manoeuvres to avoid protected zones around other air Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from
[email protected]. ECCE '14, September 01 - 03 2014, Vienna, Austria Copyright is held by the owner/author(s). Publication rights licensed to ACM. ACM 978-1-4503-2874-6/14/09...\$15.00. http://dx.doi.org/10.1145/2637248.2637264
Yakubu Ibrahim Swinburne University Hawthorn 3122, Australia
[email protected] traffic [6] without relying on the current “see and avoid” technique [7]. CDTIs that have been proposed for helping pilots avoid collisions by managing safe separation between aircraft included the geometry of relative position and velocity [8, 9, 10, 11, 12]. Chakravarthy and Ghose developed the concept of a collision cone for determining whether a collision between robots is imminent [13]. The Delft University of Technology employed the cone in their CDTI, calling it a forbidden beam. In concordance with this work, we have configured a display that has the same underlying principles of relative motion and the forbidden beam. It differs though in the expression of the vector diagram such that the higher levels of the abstraction hierarchy for an Ecological Interface Design are more likely to open to immediate perception. RELATIVE MOTION
The current practice in Air Traffic Management is to maintain a protected zone of five nautical miles of lateral separation around aircraft. Conflict is considered to occur when this zone is violated. Aircraft on a collision course must take evasive action to avert the conflict. Evasive action depends on pilots adjusting heading and/or speed such that the motion of their aircraft (ownship) relative to the intruder does not violate the protected zone. No instrument is currently available on aircraft for directly supporting pilots making collision-avoidance manoeuvres based on perceived relative motion. Consider a typical conflict scenario of two aircraft at the same altitude converging in a relative motion diagram shown in Figure 1, which is a Euclidean vector diagram of lateral separation of aircraft. It shows—using Newtonian relativity—the relative velocity of the ownship to the intruder within the bounds of the forbidden beam, thereby indicating that the aircraft are on a collision course. To avoid collision, the pilot must manoeuvre this vector out of the forbidden beam by changing the aircraft’s velocity (ground speed and heading). The ownership and intruder’s vectors are represented as shown where there is no wind. The velocity of ownership ( v own ) relative to the velocity of the intruder ( v int ) is given by: v own v own / int v int v own / int v own v int
vown vown /int vint
N Con f l i c ta ng l e
Φ
vown
vint
vint
Ow n sh ip
P ro t e c t edZ on e In t ru d e r
vown /int
F o r b id d e nB e am
F igu r e1 .Eu c l id e anv e c to rd ia g r amfo rla t e ra ls ep a r a t ion . DES IGN W ITHAPPROPR IATEAFFORDANCES
In m a in t a in i ngs a f es ep a r a t ion ,p i l o t sn e edtob ea b l et o m a n o eu v r et h e i ra i r c r a f tw i th m in im a ld ev i a t ionf romth e p l a n n e df l igh tp a th . Th a ti s ,th ea imi sto m an ag ef l i gh t p a t h sw i thc o r r e c t i on sth a tg en t lyim p a c tp a s s eng e r s ,w e l l b e f o r eac on t e n t i ona r i s e sth a tinvok e sa c t i onf romTCAS (T r a f f i can dC o l l i s ionA v o id an c eSy s t em ) . Th ei n s t rum en tm u s tw a rnap i lo to fanim p en d in gb r e a c ho f s e p a r a t i ons t a n d a rd sth roug has ig n a lth a ti ss u f f i c i en t ly s a l i e n tf o rd r aw i nga t t en t iontoth eim p en d i ngs i tu a t i onan d tos u ppo r tt h ep e r c e p t i ono fs t r a t eg i e sf o rr e so l v i ngt h e s i t u a t i on .Tor e s pondapp rop r i a t e ly ,th ep i lo tm u s tb eaw a r e o ft h eex t en to fth ev io l a t iono fcon s t r a in t sons a f ep a s s ag e. Mo r e ov e r ,f o rg o a l d i r e c t edb eh av iou r,th ep i lo tm u s ta l s o p e r c e i v eh owt oa d ju s th e ad ingan d /o rs p e edt oa l l e v i a t et h e s i t u a t i on .F o ra p i lo tw e l l -p r a c t i s edinth eu s eo fth e in s t rum e n t ,t h eb eh av iou r wou ld b esk i l lo rru l eb a s e d w i thin R a sm u s s en ’ s SRK (Sk i l l s , Ru l e s , Kn ow l e dg e ) ty po l ogy[1 4 ] .
InF igu r e3,th ef o rb i dd e nb e am ,,i ssh a d e d ,w i t hi t sa p e x l a b e l l e da s .Th ev e l o c i ty v e c t o rf o rt h e ow n sh i pi s d e p i c t e dbya na r row . Wh i l ei t sh e adi sw i t h i nth ef o rb i dd e n b e amt h ea i r c r a f ta r eonac o l l i s i onc ou r s e .Th i sd i s p l ayh a s a f f o rd a n c e st h a tp i l o t sm ayf i n dw e l l su i t e df o rop e r a t i n g u n d e rb e h a v i ou rt h a ti ssk i l lo rru l eb a s e d .Th eyc anr e ad i ly s e eth a tth eya r eona nun s a f ep a t h . Byc h ang i ngh e ad i ng a n ds p e ed ,t h em ag n i t u d ean dd i r e c t i on o ft h ev e l o c i ty v e c t o rc anb ea d ju s t e ds u c ht h a tt h eh e ado ft h ea r row m ov e sou t s i d et h ef o rb i dd e nb e am .H ow e v e r ,th ed i s p l ayin F i gu r e3 u n f o r t un a t e lyl a ck sa f f o rd a n c e st ot h eu n d e r ly in g p r i n c i p l e so fN ew t on i a nr e l a t iv i tyf o rt h ef o l l ow in gr e a s on s : 1 . Anow n sh i pp e r s p e c t i v ebyt h ep i l o ti sl o s tb e c a u s eth e p i l o t ’ sv i ew po in ti sn o ta tth ea p e xo fth ef o rb i dd e n b e am 2 . Th ei n t ru d e r ’ sp ro t e c t ed z on ei sn o t sh ow na s s u b t en d i ngt h ef o rb i dd e nb e am 3 . Th ed i s p l a yo fth ev e c t o r sf o rbo t ht h ev e l o c i tyo fth e i n t ru d e ra n dr e l a t iv ev e l o c i tyo ft h ea i r c r a f tsh ow nin F i gu r e2a r em i s s ing
F igu r e3 .D e l f tD i sp la yw i thFo rb idd enB e am[8 ] Wo rkDoma inAna lys is
F igu r e2 .D e l f tv e c to rd ia g r amfo rla t e ra ls ep a r a t ion
Th ev e c t o rd i a g r am u s e d by D e l f tin d ev e lop i ngt h e i r d i s p l ay(F i gu r e3)i ss h ow ninF igu r e2.I ti sb a s e dont h e f o l l ow i ngv e c t o rr e l a t ion s h ip ,w h i chi sar e o rd e r i n go ft h e r e l a t i on sh i ps h ow nabov e :
Th eN ew t on i a np r in c i p l e so fr e l a t i v i tyan dEu c l i d i anv e c t o r r e l a t i on sh i p s ,d i s cu s s e da bov e ,a r ea b s t r a c tp r i n c i p l e so f m o t i on r e l a t i ng c o l l i s i on av o i d a n c e a c t iv i t i e s . R e p r e s e n t a t i ono fasy s t emfo rc o l l i s i ona v o i d a n c ec anb e a n a ly s e da td i f f e r e n tl e v e l so fa b s t r a c t i on :f rom th e c om pon en t so fth e phy s i c a l sy s t emt ot h e un d e r ly in g a b s t r a c tp r i n c i p l e s . Wo rkD om a inA n a ly s i s(WDA )i sat em p l a t e ,d e v e l op e dby R a sm u s s en ,f o ra n a ly s ing an dm od e l l in gsy s t em su s in g v a r i ou sl ev e l so fa b s t r a c t i on[1 4 ].I ti sag e n e r i cf r am ew o rk f o rd e s c r i b i ngg o a l o r i en t e dsy s t em s ,ina w ayt h a td e p i c t s t h e i rpu rpo s i v ea n dphy s i c a la s p e c t s[1 5 ,16 ,17 ] .I ti son e
of the two types of analysis incorporated in Cognitive Work Analysis (CWA). In that it describes the system in an event independent way, WDA is quite separate from Cognitive Task Analysis (CTA), which is an event dependent analysis of the activities that take place within the domain. That is, CTA represents the activities of decision agents in interaction with a system described by WDA. WDA uses an Abstraction Hierarchy that comprehensively describes a system at each level of abstraction. The higher levels describe the reasons or purposes of the physical system described at the lower levels [18]. Rasmussen called the levels—from the most to least abstract—functional purposes, abstract functions, generalised functions, physical functions and physical form (Figure 3). Since the framework was devised in the 1980s, improved terminology has evolved that denotes the levels unequivocally. They are—in like order—functional purposes, values and priority measures, purpose-related functions, object-related processes and physical objects [19]. The uppermost level of the abstraction hierarchy expresses the functional purposes of the system. Levels are linked through “means-ends” relationships; arcs from nodes at each level link nodes at the level below, such that the lower-level nodes represent the means for obtaining the ends for the upper-linked nodes. The level of abstraction that is applicable for the description of the system and its constraints varies with activities associated with CTA, which may be skill, rule or knowledge based. The form that is appropriate for displaying data varies with the type of decision being made. The presentation and control of information between the system and human may vary with the recognition-action cycles of the cognitive tasks. The tasks associated with recognising the state of the system are linked to the abstraction hierarchy of the work domain. To support behaviour within the SRK typology, goalrelevant constraints from the work domain should map onto salient perceptual properties of the display, at the various levels of abstraction. A display that represents concurrently the various levels of abstraction, Vicente and Rasmussen called an Ecological Interface Design (EID) [20]. Its goal is to allow activities to be performed at the level of cognitive demands that the task requires. If constraints are mapped into an interface that is based on an abstraction hierarchy, pilots’ mental models can be captured and externalised to improve situation awareness [21]. Such a configural display, which can be directly manipulated by the user, may help novice pilots form a mental model that properly represents the system. It can act as externalised form of the user’s mental model, thereby supporting activities that would otherwise be held in working memory [22]. FLIGHT COLLISION AVOIDANCE SYSTEM
The functional purpose of the Flight Collision Avoidance System (FCAS) is to assist pilots to manage the safe
separation from other air traffic as the flight progresses. The level of functional purposes in the abstraction hierarchy pertains to the purpose and constraints affecting the system’s operation. These processes are influenced by external constraints that are based on standard rules, laws or tests for achieving the main objective or purposes of the system. A goal of pilots operating in a free flight environment is not to violate separation constraints with other aircraft. Therefore, the functional purposes of a FCAS that supports activities associated with this goal are depicted at the top level of the abstraction hierarchy shown in Figure 4 as:
Detect potential violation of separation
Provide resolution guidance
A display based on EID allows pilots to access information at the level of abstraction they require when making decisions. Typical questions may be: How close is their aircraft to the intruder’s protected zone? How much faster or slower should they fly to avoid collision without changing heading? What change in heading, while maintaining speed, should they fly to avert conflict? These questions concern relative relationships between aircraft, which are derived from information from physical sensors. That is, their observation is at the level of purpose-related functions. The principles and performance metrics that direct the processes underlying the system’s progression towards its functional purposes is at the level of values and priority measures. The laws governing conflict zones are defined and controlled by minimum separation standards, while performance is judged by suitable metrics—concerning, perhaps, increase in flight path, flight time or fuel burnt— and observation of inviolable constraints, such as not transgressing the boundaries of the protective zones. Below the level of values and priority measures is the level of purpose-related functions. This level is in terms of the functions, responsibilities and processes that influence the constraints at the level above. It provides pilots with necessary information for performing flying activities such as lateral manoeuvres to avoid obstacles. Pilots are required to manoeuvre their aircraft outside the intruder’s protective boundary, i.e., not violating minimum separation standards. Fundamental physical quantities of aircraft are relative positions, velocity and protected zone. There are key elements of conflict geometry. The ability of pilots to change aircraft direction is constrained by aircraft manoeuvrability. An aircraft’s manoeuvring capability for lateral control is also constrained by the minimum separation standard and wind speed. However, understanding these constraints provides pilots with the mechanism to control and maintain the desired flight path.
Functional Purposes
Values & Priority Measures (Abstract Functions)
Purpose-Related Functions (Generalised Functions)
Object-Related Processes
(Physical Functions)
Physical Objects (Physical Form)
Provide resolution guidance
Detect potential violation of separation
Physical Laws of Collision Avoidance
Protective zone
Ownship vector relationship between ground velocity, air speed & heading
Ownship heading
Ownship heading sensor
Ownship position relative to intruder
Ownship air speed
Ownship airspeed sensor
Ownship position
Ownship position sensor
Ownship velocity relative to intruder
Ownship ground velocity
Ownship velocity sensor
Intruder position
Intruder position sensor
Intruder ground velocity
Intruder velocity sensor
Figure 4. Abstraction hierarchy for FCAS Formation of the FCAS display
At the second lowest level of the hierarchy is detailed information about the physical system in terms of objectrelated processes: the means for achieving the functional purpose of the system. It enables pilots to interact with the system at the physical functional level. The lowest level specifies the actual physical objects for the functions described by the object-related processes: for example, the physical object for obtaining the function of airspeed is depicted in Figure 4 with a generic label of airspeed sensor. In its simplest form, this instrument displays indicated airspeed, which is not true airspeed as it is uncorrected for changes in temperature and altitude. In the FCAS, ground velocity is derived from a combination of sources including GPS and the on-board inertial navigation system.
The display for the FCAS was developed as shown in Figure 5 by starting with the Euclidean vector diagram for lateral separation (Figure 1). It is a presentation of the FCAS at the level of purposed-related functions in the abstraction hierarchy (Figure 4). The circle around each aircraft shows the constraint boundary of five nautical miles for the protective zone that must not be violated. The vector labelled “e”, depicting the velocity of the ownship relative to the intruder, when inside the forbidden beam signifies that the aircraft are on a course for which separation will be violated. That is, a view of the FCAS at the level of value and priority measures is accessed via the protected zone, the forbidden beam and vector relationships.
Figure 7. FCAS without EID. Figure 5. FCAS addition to the primary flight display.
As aircraft move closer, the white line labelled “f” denotes loss of separation. Accompanying the loss of separation, the angle of the forbidden beam increasing, thereby adding salience to the signalling of its rate of loss.
The experiment was conducted on a standard desktop computer and colour monitor (using 1024 x 768 XGA with a graphics card) with the inclusion of a Saitek Pro Flight System. The simulator software and the interface for the yoke and throttle were written in C++ language and MATLAB script. The software enabled pilots to avoid obstacles by tracking, navigating, maintaining or deviating from the intended flight path. The algorithm was a levelaircraft conflict resolution of flying a twin-engine aircraft under no wind conditions [9]. Aircraft dynamics were not modelled. Participants were asked to perform tasks under simulated Instrumental Flight Rules. Each experimental run consisted of three conflict scenarios—head on, port and starboard approaches—in which a conflict occurs. Both the ordering of the scenarios and the heading for each scenario were randomly assigned. The ownship was allowed to manoeuvre to avoid conflict, while the intruder maintained a constant heading that was randomly assigned.
Figure 6. FCAS based on EID.
The image of a miniature aircraft is centred on the navigational compass. The arc labelled “j” at the top of the display represents the bank angle range. This guides the pilot to not violate physical parameters of flight (which would be included in a separate abstraction hierarchy with flight as the functional purpose). EXPERIMENTAL TEST OF FCAS DISPLAYS
Under simulated flight conditions, two collision avoidance displays—one FCAS based on EID (Figure 6) and the other not (Figure 7)—were studied under experimental conditions. The twenty one participants recruited were either aviation students or professional pilots, of which 81% were between 18 and 40 years old and the others were older than 50 years. Divided into experimental (13) and control (8) groups, they used the EID and non EID displays, respectively. The experiment lasted approximately an hour.
Participants were instructed to fly a specific flight route by an ATC command shown in the upper right corner of the display. They continued to fly the prescribed flight route and airspeed with minimum deviation as commanded by ATC. Each scenario began with a different flight course. Participants observed the development of conflict scenario for seventy seconds. There was one or two other aircraft, which could pose a threat to the ownship protected zone. The FCAS display provides participants with basic information about others aircraft relative positions, velocity vector, altitude and airspeed of a nearby aircraft. The forbidden beam appeared when the intruder position was 21 nautical miles relative to the ownship position. Participants were required to start manoeuvring only on a TCAS-like audio warning of a threat. A standard rate of turn (three degrees per second) was used for a heading change. A rate of change in aircraft heading was a function of bank angle. Both the control and experimental FCAS incorporated the
arc for displaying bank angle, as discussed above. Pilots had a numeric display of ground speed on the FCAS. Post-trial questionnaire
To evaluate the system’s usability, participants undertook a post-trial questionnaire that used a five-point Likert scale for the questions shown in Table 1. The results in Figure 8 show that for all questions, the users of the EID display responded more favourably than those using the non EID version. The results for the questions (Q2, Q4 and Q6), which relate directly to the primary purpose of the FCAS, indicate that while users found the EID display somewhat more useful than the non EID version for locating intruders an avoiding conflict, the EID form of the FCAS was perceived to be much more useful for following a desired flight path. Q1
Figure 8: Mean ratings on seven subjective preferences questions for experimental and control group.
The response of the yoke and throttle was too sensitive for me to track the path I wanted to follow
Q2
How useful was the system for understanding the location of the Intruder relative to the Ownship?
Q3
How useful was the system for understanding the direction of travel of the Intruder?
Q4
How useful was the system for avoiding conflict?
Q5
How useful was the system for avoiding stall when manoeuvring the aircraft?
Q6
How useful was the system for providing sufficient information for following a desired flight path?
Q7
Please rate your overall opinion of the system. Table 1: Post-trial questionnaire
Correlation analyses were conducted on the responses to the questions, to determine how well the dependent variable correlates with the participants’ ratings. Results of the analyses are plotted in Figure 9 and Figure 10. Participant pilots are shown as nodes; arcs represent the correlation between paired pilots. Correlation coefficients of 0.7, 0.8 and 0.9 are shown as light blue continuous lines, black dashed lines and black continuous lines, respectively.
Figure 9: Spider-web visualization of the correlation matrix for the experimental group.
The patently obvious difference between pairing for the two displays reflects a difference in responses to the questions. The common pattern of responses by users of the EID display is not matched by the users of the non EID version. These results indicate—prima facie—that pilots using the non EID FCAS had very mixed assessments of the display. A “positive” conclusion that is tempting is to conclude that a design of an FCAS based on EID principles is superior. However, a scientifically sceptical view is that the display used as a control is that of a “straw man.” The problem that the researchers encountered was that flight collision avoidance systems used by pilots for maintaining spatial separation are not currently available. Therefore, they had
to develop a non EID display that represents all information for the separation task in a “conventional” way.
REFERENCES
1. Sheridan, T.B. Human Factors Research Needs for NextGen-Airportal Safety. NASA/CR–2009-21537, (2009). 2. Metzger, U. and Parasuraman R. The role of air traffic controller in future air traffic management: An empirical study of active control versus passive control. Human Factors 43, 4 (2001), 519-528. 3. RTCA. Final report of the RTCA board of directors' select committee on free flight. RTCA, Inc., Washington, DC, USA, 1995. 4. Hoekstra, J. M., van Gent, R. N. H. W., & Ruigrok, R. C. J. (2002). Designing for safety: the ‘free flight’ air traffic management concept. Reliability Engineering & System Safety, 75(2), 215-232. 5. Hilburn, B., Parasuraman, R., Jha, P. and McGarry, K. Emerging human factors issues in future air traffic management. Center for Human Performance Research. (March 2006) http://www.chpr.nl/
Figure 10: Spider-web visualization of the correlation matrix for the control group. CONCLUSION
The design of the flight collision avoidance system (FCAS) was based on the principles of Ecological Interface Design (EID), with the objective of making constraints and principles of relative motion visible to pilots. The FCAS with EID has two major elements associated with purposerelated functions and measures of values and priority: (a) A forbidden beam display (b) A relative motion display
6. Thomas, L.C. and Rantanen, E.M. Human factors issues in implementation of advanced aviation technologies: a case of false alerts and cockpit displays of traffic information. Theoretical Issues in Ergonomics Science 7, 5 (2006), 501-523. 7. Zingale, D. and Williems, B. Review of Aircraft SelfSpacing Concepts: Implications for Controller Display Requirements. Federal Aviation Administration. DOT/FAA/TC-TN-09/03 (2009). 8. Van Dam, S.B., Mulder, M. and van Paassen, M.M. Ecological interface design of a tactical airborne separation assistance tool. IEEE Transactions on Systems, Man and Cybernetics, Part A: Systems and Humans 38, 6 (2008), 1221-1233. 9. Bach, R., Farrell, C. and Erzberger, H. An Algorithm for Level-Aircraft Conflict Resolution. NASA/CR-2009214573 (2009).
Previous developed experimental collision avoidance systems based on EID inadequately map the relationship between these components, as they do not clearly show the geometry of conflict and operational constraints. With wellmapped constraints, pilots may be able to enhance their mental models of collision avoidance and then make use of the mental model that is externalised in the FCAS to improve situation awareness.
10.Andrews, J.W. A relative motion analysis of horizontal collision avoidance. Proceedings of the 15th Annual Symposium, Las Vegas, SAFE Association (1977).
While experimental results are inconclusive, they show that there were significance differences in the behaviour of the users of EID and non EID displays. It seems that what is important in collision avoiding system’s design for a free flight environment is that they should support pilots in guiding them in such a way that their performance is consistent and safety is not compromised.
12.Bilimoria, K.D. A geometric optimization approach to aircraft conflict resolution. In Proc. AIAA Guidance, Navigation, and Control Conference and Exhibit, Denver, CO (2000).
11.Xu, X. and Rantanen, E.M. Effects of air traffic geometry on pilots’ conflict detection with cockpit display of traffic information. Human Factors 49, 3 (2007), 358-375.
13.Chakravarthy, A., and Ghose, D. (1998). Obstacle avoidance in a dynamic environment: A collision cone approach. IEEE Transactions on Systems, Man and Cybernetics, Part A: Systems and Humans28, 5 (1998), 562-574.
14.Rasmussen, J. Skills, rules, and knowledge; signals, signs, and symbols, and other distinctions in human performance models. IEEE Transactions on Systems, Man and Cybernetics SMC-13, 3 (1983), 257-266.
19.Naikar, N., Hopcroft, R. and Moylan, A. Work domain analysis: Theoretical concepts and methodology. Air Operations Division, DSTO, DSTO-TR-1665, Australia, 2005.
15.Rasmussen, J., Petjersen, A.M. and Goodstein, L.P. Cognitive systems engineering. John Wiley, New York, 1994.
20.Vicente, K.J. and Rasmussen, J. Ecological interface design: theoretical foundations. IEEE Transactions on Systems, Man and Cybernetics, 22, 1992, 589-606,
16.Vicente, K.J. Cognitive Work Analysis: Towards Safe, Productive, and Healthy Computer-based Work. Lawrence Erlbaum Associates, Hillsdale, NJ. 1999.
21.Amelink, M.H.J., van Paassen, M.M., Mulder, M. and Flach, J.M. Total Energy-Based Perspective Flight Path Display for Aircraft Guidance along Complex Approach Trajectories. In Proc. Twelfth International Symposium on Aviation Psychology, Dayton, OH, 2003 .
17.Higgins, P.G. Architecture and Interface Aspects of Scheduling Decision Support. In B. MacCarthy and John Wilson (Eds.) Human Performance in Planning and Scheduling. Taylor and Francis, (2001), 245-281. 18.Rasmussen, J. The role of hierarchical knowledge representation in decisionmaking and system management. IEEE Transactions on Systems, Man and Cybernetics SMC-15, 2 (1985), 234-243.
22. Rasmussen, J. A cognitive engineering approach to the modeling of decision making and its organization in process control, emergency management, cad/cam, office systems, and library systems. In W.B. Rouse (Ed.) Advances in Man-Machine Systems Research, JAI Press, Connecticut, Vol. 4, 1986, 165-243.
