Contemporary Issues in Human Factors and Aviation Safety _09!05!2010

|
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
 13 views
of 14

Please download to get full document.

View again

Description
CONTEMPORARY ISSUES IN HUMAN FACTORS AND AVIATION SAFETY Edited by: Don Harris – Helen C. Muir 1 FEASIBILITY OF APPLICATION OF HCA PRINCIPLES The previously described HCA principles may sound useful, but the real test is whether they can be used in practical design environments. This section therefore firstly gives two short examples of the utilization of some of the principles for real future ATM design projects, focusing on computerized support tools (automation) for controllers, and based on
Share
Tags
Transcript
  CONTEMPORARY ISSUES IN HUMAN FACTORS AND AVIATION SAFETYEdited by: Don Harris – Helen C. Muir 1 FEASIBILITY OF APPLICATION OF HCA PRINCIPLES The previously described HCA principles may sound useful, but the real test iswhether they can be used in practical design environments. This section thereforefirstly gives two short examples of the utilization of some of the principles for realfuture ATM design projects, focusing on computerized support tools (automation) forcontrollers, and based on UK projects. It then gives a most extensive account of astudy using a ‘deeper’ HCA approach to develop conflict resolution tool controllers.  Application to future trajectory prediction and medium term conflict detectiontools for En Route airspace. The operational concept for these tools, under development (Whysall, 1998) is to’provide genuine support to the current controller roles, rather than to change thoseroles significantly’. This indicates that the tool-set is designed around the waycontrollers currently work and will work in the future. The toll do not change theway controller operate, but provide additional facilities to allow them to do whatthey currently do more easily. From an HCA viewpoint, the tools have been designedto allow controllers to continue their decision-making activities, but to do so moreefficiently. For example, a Co-ordination List tool would help the Planner to decidewhen to co-ordinate aircraft into or out of the sector. This supports the Planner’stime management.When the controllers have decided to perform a co-ordination, lateral and verticalrisk tools allow them to judge better what other aircraft are likely to interact withthe aircraft at or beyond the sector boundary. Such tools present a graphicalrepresentation of the projected positions of both the subject aircraft and those inthe vicinity. The controllers currently use their mental picture to mentally gauge theposition of aircraft. The fact that is based on the controller’s judgment means that itis influenced by factors such as experience. Consequently, controllers will report thatthey sub-consciously accompany each mental projection with an idea of uncertainty(usually described as a ‘gut feeling’). The tools being designed calculate theuncertainty within future projections and graphically indicate this to be controller,helping them to ‘distinguish’ between acceptable, uncertain and unacceptablesituations’. This is therefore an example of where the tools are built around thecontroller’s mental models for controlling traffic.Electronic stripsIn a number of ATM organizations there is a trend towards electronic strips, ratherthan paper flight strips as are currently used (see also Hopkin, 1995). When designingfor electronic strip systems in the UK, it was considered that flight strips are acommon thread between the activities of the Planner and Tactical controllers. Theyare said to “provide an element of ‘team’ focus for the controllers’ (gooddship andPember, 1998).This important benefit ism one of the key principles that electronic  CONTEMPORARY ISSUES IN HUMAN FACTORS AND AVIATION SAFETYEdited by: Don Harris – Helen C. Muir 2strips seek to maintain. Furthermore, is seeks to perpetuate other human-centredprinciples: ã The design of the flight strips and the strip boards will be flexible enough ‘toallow controllers to use the information presented to assimilate the airspacepicture to their individual preference’. The system will permit controllers toorganize their strips in a way that continues to support the mentalrepresentation of the picture. ã Recording information on electronic strips necessarily moves away frompennon paper but aims to be ‘simple and quick’. The designers recognize that‘the entry of flight data becomes second nature and does not distract thecontroller from the primary task of traffic separation’. This has been achievedby minimal keyboard use and easily accessed strip fields, and using userinterface support aspects such as ‘implicit focus’. ã The electronic strips do not appear to increase the controllers’ workload,either in terms of ascertaining flight data or manipulating (changing) it. Thesystem has been designed so that the strip information has been divided intothree types. Thus, information controllers need all the time is alwaysavailable; the information they need some of the time (but when they do theyneed permanently displayed for an aircraft) is instantly accessible; theinformation controllers need only occasionally can be revealed easily and thenremoved automatically to reduce clutter. ã The controllers are supported in their conflict detection by electronic strips,as it has a facility to highlight aircraft likely to come into conflict in the nearfuture, based on the same flight plan information that controllers normallyuse. ã The electronic strip concept provides a useful link that did not exist beforebetween the strip and the radar display. When the controller selects anelectronic strip, the track data block for that aircraft is also highlighted onthe radar. This is likely to help controllers to integrate their information andhighlight errors. ã The prototype system has recently been analyzed for human error potentialusing a human Hazard and Operability approach (Kennedy et al, 2000). Thisstudy identified errors via a group error identification approach, consideredtheir consequences, and determined design changes that would prevent theerrors or facilitate detection and recovery. ã The system, and the previous toolset in (i) above, as they develop, are beinganalyzed in prototyping and real-time simulations, and workload is beingevaluated throughout its development cycle using a range of workloadassessment tools (Atkinson et al, 1997; Kirvan et al, 1998b).  CONTEMPORARY ISSUES IN HUMAN FACTORS AND AVIATION SAFETYEdited by: Don Harris – Helen C. Muir 3The tools reviewed briefly above are still under development, and so acomprehensive comparison against the principles in this report is neither appropriatenor possible at this stage. Nevertheless, there is clearly already some serious uptakeof HCA within the community designing prototype ATM system and tools. Theprogressive adoption, adaptation and application of the HCA principles contained inthis paper should further ensure that the automation complements, rather thancompetes, with the controller, and that the requerid safe increase in capacity isachieved. The example below shows a more formal and analytical approach to thedevelopment of a toll that will match the controller’s mental model of a critical taskin ATM, namely aircraft conflate resolution. CORA – human centered automation inthe development of a conflict resolution assistant (CORA) system. A European study related to Human Centered Automation called the RHEA project(Role of the Human in the Evolution of ATM: Nijhuis, 2000) considered futureautomation prospects and their likely impact on the controller. The study tried todetermine what level of automation would result in best performance. Differentlevels were considered, from full automation to fully manual, but with many levels inbetween, such as computer-generatd advice, with the controller making thedecision, to the computer implementing its own decision unless the controller vetoedit. There were also some flexible levels of automation (e.g the automation takingover when the controller’s workload became too high, or when the controllerrequested it. ) The different levels were evaluated qualitatively, and also apredictive error analysis was carried out to try to determine the best level of automation (Kirwan, 2001). The results were that the best levels evolved thecomputer or machine giving advice, and then controller deciding to accept or rejectit. Furthermore, one contrition favored particularly well at this level of automation,was called ‘cognitive tools’. The concept of a cognitive tools is that the toll itself,which gives advice to the controller, is derived around the controller’s own mentalmodel of how the situation should be resolved, as opposed to being derived frompurely mathematical models etc. Such an approach can be seen as a specific form of ‘Human Centred Automation’ (Billing, op cit). The Euro control conflict ResolutionAssistant (CORA) Project aims therefore to produce a controller-centrad approach toconflict resolution, using the !Cognitive Tools’ concept. Currently, controllers aremasters at real-time conflict detection and resolution, and this expertise in theseparticular system functions is the result of rigorous selection and intensive training inair traffic control over a prolonged period. Conflict detection and resolution areindeed seen as core functions of the controller today, i.e. controller, when asked todefine their job simply, often say ’separating aircraft’ . Any tools that thereforepurport to support such functions have two main obstacles to overcome. The first isthe development and provision of a viable alternative that is at least as goodcontroller expertise (and preferably better). The second is ensuring that such toolswill l be used by controllers, when those very tools can be seen as a threat to thosesame controllers. This latter aspect is poignant , since conflict resolution is seen as  CONTEMPORARY ISSUES IN HUMAN FACTORS AND AVIATION SAFETYEdited by: Don Harris – Helen C. Muir 4core task and skill of the controller, especially given their responsibility andculpability should separation be lost. Today, assistance whit conflict detection existsin many places via various forms of short term conflict alert, which was of impendingloss of separation (e.g. a minute or so) and in several air traffic centers, mediumterm conflict detection is now being piloted and implemented (a – 15 minuteswarning). Therefore, assuming medium term conflict detection systems area a leastmoderately successful (i.e they enhance air traffic management and are usedeffectively by the controllers), the next step to consider is conflict resolution. Aconflict resolution tool would be able to advise controller in terms of which way toturn conflicting aircraft, or whether to climb or descend one or both aircraft est.Furthermore, such resolution advise could be based not only on safety parameters,but operational ones also, e.g turning one aircraft out of a conflict, but giving it aturn which takes it the least distance away from its optional route. This is effectivelywhat controllers do now, and they do it well. However, in the future with moretraffic and more varied traffic patterns, a support toll could improve matters andhelp maintain safety and quality of service to the airlines and passengers.There is therefore a need a tool to help conflict resolution. Conflict resolutionis in fact a difficult area. There are many potential approaches to conflict resolution(e.g. see Kuchar and Young, 1999; Mendoza, 1999; Kirwan 2002a), many of which canwork in theory. However, none have been proven in practice, although one is nearingcompletion (Kirk et al, 2000). Several of the models are mathematical in nature.Such models consider the conflict geometry and the surrounding airspace and simplyplot a way out that maintains minimum separation. However, such ‘clinically correct’resolution may be highly inefficient for the aircraft involved. The controller, incontrast, is able to resolve the conflict in a way that minimizes penalties to aircraft,and exhibits a certain degree of ‘fairness’ to the aircraft involved. This isappreciated by the airlines. This ability to optimize according to a range of parameters, in situations that are almost infinite in their variety, is somethinghumans are surprisingly good at, due to their ability to make trade-offs and theirpattern recognition faculties.The area of conflict resolution is therefore complex. A large part of thiscomplexity is due to the multi-dimensionality of the area, in that it is not simplyabout aircraft conflict geometry, but also about optimization according to a range of criteria, in a fairly open system environment subject to significant traffic variations(e.g. peaks and troughs), weather effects, etc. Given that the controller currentlydoes this job well, and that automation tools are as yet unproven, it would seemwise to start with a tool that can assist the controller. In fact, what is needed issome automation to help the controller, but the controller must remains in charge,and must be able to decide if the advice the toll is giving is reasonable and can beaccepted as it is, or if it needs to be adapted due to some local circumstances. Thisrequirement indeed sounds like a request for Human Centred Automation.
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks