IT for Error Remediation And Trapping Emergencies ... · PDF fileModelling in Assisted...

1

ITERATE Special Issue - October 2010

Created by

The research leading to these results has received funding from the European Commission’s Seventh Framework

(FP7/2007-2013) under grant agreement n. 218496

Project ITERATE

IT for Error Remediation And Trapping Emergencies

NNNNNNNNEWSLETTEREWSLETTEREWSLETTEREWSLETTEREWSLETTEREWSLETTEREWSLETTEREWSLETTER

1.HMAT - 2010 WORKSHOP 1

2.ABSTRACTS OF KEYNOTES 2 - 5

3.ABSTRACTS OF ITERATE PAPERS

6-7

5.TABLE OF PUBLICATIONS 9 - 10

4.SIMULATOR 8

In this issue: 1. HMAT 1. HMAT 1. HMAT 1. HMAT ---- 2010 W2010 W2010 W2010 WORKSHOPORKSHOPORKSHOPORKSHOP

Special Thanks:

We wish to thank all HMAT speakers, particularly the keynote speakers and the

session chairpersons as well as all participants to the

conference

The ITERATE project has recently sponsored, with two other European projects, namely ISi-PADAS and HUMAN, the HMAT-2010 (Human Modelling in Assisted Transportation) Workshop, which was held in Belgirate from 30 June to 2 July and was at-tended by about 65 participants from all over the world. The conference dealt with the human modelling into design processes and in safety assess-ments of innovative technologies in highly assisted systems. The aim of such models is to provide an improved understanding of the human factors and to predict performance and behaviour of the human in interaction with new technologies in normal and emergency situations, for all surface transport modes and for cockpit environments. Renowned speakers from Europe and through the world participated to the Workshop. HMAT organizers were particularly happy to welcome as key-note speakers: Erik Hollnagel, from MiNES ParisTech (France); Toshiyuki Inagaki, from the UT – University of Tsukuba (Japan); Nick McDonald, from TCD - Trinity College of Dublin (EIRE); Andrew M. Liu, from MIT - Massachusetts Institute of Technology (USA); Truls Vaa, from TØI - Institute of Transport Economics (Norway);

-ATE; Pietro Carlo Cacciabue and Marc Vollrath for the ISI-PADAS and finally Andreas Luedtke and Denis Javaux, for HUMAN. The HMAT-2010 Workshop offered the opportunity to have fruitful scienti-fic discussions on three main topics: • Advanced human models in tran-

sportation • Human Errors and Risk Asses-

sment in design processes of assi-stance systems

• Methods and tools to prevent erroneous behaviour to mitigate its consequences.

The book containing proceedings is in progress and it will be published by Springer. During the HMAT-2010 Workshop, ITERATE partners gave the possibility to all participants to visit the car and train simulator utilised inside the pro-ject for the empirical experiments. In the following pages, you can find: • The abstracts of the papers presen-

ted during the conference by Keynotes and by ITERATE Par-tners;

• The presentation of the simulator utilised for ITERATE experiments.

• The complete table of publications by partners associated to the work of the ITERATE project.

Figure 1. HMAT Participants

Klaus Bengler, from TUM - technical Uni-versity of Munich (Germany) and Brian Gore, from NASA - National Aeronautics and Space Administra-tion (USA). Other Keynote speakers presented the three sponsor Projects: Mag-nus Hjälmdahl, Oliver Carsten and David Shinar for the ITER-

2

ITERATE Special Issue- October 2010

TTTTHEHEHEHE HUMANHUMANHUMANHUMAN ININININ CONTROLCONTROLCONTROLCONTROL : M: M: M: MODELLINGODELLINGODELLINGODELLING WHATWHATWHATWHAT GOESGOESGOESGOES RIGHTRIGHTRIGHTRIGHT VERSUSVERSUSVERSUSVERSUS MODELLINGMODELLINGMODELLINGMODELLING WHATWHATWHATWHAT GOESGOESGOESGOES WRONGWRONGWRONGWRONG

that are be completely specified, it is not reasonable for systems that are underspecified. Since this latter category includes most of the socio-technical systems we have to deal with in today’s world, a different approach is required. Instead of looking at joint system perform-ance as either right or wrong, it should recognise that coping with complexity means that perform-

ance necessarily must be variable in order to compensate for the un-derspecification of work and ac-tivities. Models and methods must therefore be able account for that.

The study of human-machine systems or joint cognitive systems has traditionally tried to describe and model what the system – and therefore also the humans – should do. When systems performance differed from design specifica-tions, it was explained as a fail-ure of either the technology or of the humans. While this approach might be reasonable for systems

Erik Hollnagel

MINES ParisTech—Sophia Antipolis, France [email protected]

TTTTHEHEHEHE AAAARTRTRTRT TOTOTOTO MAKEMAKEMAKEMAKE ANANANAN EEEERRORRRORRRORRROR

TTTTHEHEHEHE DDDD ILEMMAILEMMAILEMMAILEMMA BETWEENBETWEENBETWEENBETWEEN PPPPREVENTIONREVENTIONREVENTIONREVENTION , L, L, L, LEARNINGEARNINGEARNINGEARNING ANDANDANDAND MMMM IT IGATIONIT IGATIONIT IGATIONIT IGATION

Klaus Bengler

Institute of Ergonomics, Technische Universität München Boltzmannstr. 15, 85747, Garching, Germany

[email protected] question on avoidable consequences is more suitable.On the other hand erroneus behavior is a source of infor-mation and learning for human opera-tors .

The contribution will give examples for the dilemma of error free environ-

ments and question how the trade-off between erroneous and learnable could be used proactive.

Following the established tradition of user centered system design leads to the effect that erroneous behavior of human operator and technical system shall be minimized.

As this development goal is in most system constellations to advanced the

MMMMODELL INGODELL INGODELL INGODELL ING DDDD IFFERENCESIFFERENCESIFFERENCESIFFERENCES ININININ BBBBEHAVIOREHAVIOREHAVIOREHAVIOR WWWW I THINITHINITHINITHIN ANDANDANDAND BBBBETWEENETWEENETWEENETWEEN DDDDRIVERSRIVERSRIVERSRIVERS

Andrew M. Liu

Man Vehicle Laboratory, Dept. Of Aeronautics & Astronautics, Massachusetts Institute of Technology, 77 Massachusetts Ave, 37-219, Cambridge, MA 02139 USA, [email protected]

This “smart car” would observe the driver’s pattern of behaviour in terms of their control of the vehicle then infer their current driving task using a Markov Dynamic Model. The approach could recognize driver actions from their initial behaviour with high accuracy un-der simulated driving conditions. Since that time new computational approaches and improved in-vehicle technology (e.g., GPS technology, advanced radar and video/computer vision, etc) have moved the realization of this concept further along. Yet, one fundamental question still needs

to be carefully addressed: Can these driver models, built on statis-tical descriptions of driver behav-iour, accurately model the differ-ences between drivers or changes within an individual driver’s be-haviour? In this paper, I describe some examples these differences and discuss their potential impact on a model’s ability to consistently recognize behaviour. To ensure the acceptance of the next generation driver assistance systems, these issues will have to be resolved.

A new generation of driver assis-tance systems such as advanced collision warning and intelligent brake assist are now available op-tions for the modern automobile. However, the addition of each new system increases the information load on the driver and potentially detracts from their ability to safely operate the vehicle. Over 10 years ago, we suggested that a car that could infer the current intent of the driver would be able to appropri-ately manage the suite of systems and provide task relevant informa-tion to the driver in a timely fashion.

2. ABSTRACTS OF KEYNOTES2. ABSTRACTS OF KEYNOTES2. ABSTRACTS OF KEYNOTES2. ABSTRACTS OF KEYNOTES

Cacciabue, P.C.; Hjälmdahl, M.; Luedtke, A.; Riccioli, C. (Eds.) 2011. Human Modelling in Assisted Transportation - Models, Tools and Risk Methods. ISBN: 978-88-470-1820-4, with permission from Springer, Milan

3


DDDDRIVERSRIVERSRIVERSRIVERS ’ ’ ’ ’ INFORMATIONINFORMATIONINFORMATIONINFORMATION PROCESS INGPROCESS INGPROCESS INGPROCESS ING , , , , DECIS IONDECIS IONDECIS IONDECIS ION ----MAKINGMAKINGMAKINGMAKING ANDANDANDAND THETHETHETHE ROLEROLEROLEROLE OFOFOFOF EMOTIONSEMOTIONSEMOTIONSEMOTIONS : : : :

PPPPREDICT IONSREDICT IONSREDICT IONSREDICT IONS OFOFOFOF THETHETHETHE RRRR ISKISKISKISK MMMMONITORONITORONITORONITOR MMMMODELODELODELODEL

problem statements about accident causation as relative risks allow you to compare risk levels of different road conditions, road user activities, an driver states. The paper objectives are fourfold: 1) To see if ITS address ma-jor traffic safety problems, 2) Provide an empirical base for discussing ef-fects of ITS theoretically, 3) Discuss the role of emotions in drivers’ infor-mation processing, and 4) Link rela-tive risks and ITS to predictions based on a model of driver behaviour – which in this context would be the Risk Monitor Model (RMM). A list of 21 relative risks is provided and for each of them an indication of whether there exist an ITS that might mitigate the problem is stated. Based on the list of known relative risks, the relevance of paradigms pro-

vided by evolution and neuroscience is suggested. The RMM, which is an eclectic model based and assembled on several older, established models and theories is elaborated and described. The major contributions to the RMM are: Näätänen and Summala’s “Zero-Risk Model (1974), Antonio R. Damasio’s neurobiological model elaborated in his book ”Descartes’ Error: Emotion, Reason, and the Human Brain” (1994), Taylor’s paper on GSR (1964), Bechara et als’ paper on SCR (1997) Wilde’s RHT (1982) and Ulleberg’s paper on personality sub-types of drivers (2002). Finally, predictions of the RMM about the outcome of ITS are stated as seven specific hypotheses.

Truls Vaa

Institute of Transport Economics, Gaustadalléen 21, N-0349 OSLO, Norway, [email protected]

The interaction between motivational forces, attention, risk monitoring and choice of driving speed is complex and continues to be a matter of con-cern, not least because of risk compen-sation and behavioural adaptation seen by the introduction of new traffic safety measures. A special matter of concern are Intelligent Transport Sys-tems (ITS), where driver behaviour models differ in their ability and ambi-tion to predict effects of ITS. The present paper discusses issues of perception, distraction, unconscious and conscious routes of information processing and decision-making and three major topics are addressed: Rela-tive risks, risk monitoring, and Intelli-gent Transport Systems (ITS). A consideration of relative risks is proposed as a fruitful angle to draw up

OOOOPERATIONALPERATIONALPERATIONALPERATIONAL MMMMODELLINGODELLINGODELLINGODELLING ANDANDANDAND DDDDATAATAATAATA IIIINTEGRATIONNTEGRATIONNTEGRATIONNTEGRATION FORFORFORFOR MMMMANAGEMENTANAGEMENTANAGEMENTANAGEMENT ANDANDANDAND DDDDES IGNES IGNES IGNES IGN

Nick McDonald et al.

Aerospace Psychology Research Group School of Psychology Trinity College Dublin, Ireland [email protected]

Increasing focus on managing the performance of operational sys-tems is driven by the relentless need to improve efficiency and save cost, by new safety manage-ment regulation and by growing interest from manufacturers in de-sign for operability. There is increasing convergence in the needs of these activities, which have traditionally been done inde-pendently. Addressing these challenges in the design and management of opera-tional systems depends on having a model of how such systems work. Such a model should be supported by data. While an enormous

amount of data is generated by operational systems, too often much of that data is unused, lacks integration or is not accessible for use. A new framework for modeling

operational systems in aviation links operational models to smarter data integration support better operational management, better design capabilities and more effective regulation.


Figure 2. ...Some Participants visiting the ITERATE simulator...

4


An error in situational recognition may occur while driving a car, and the error can sometimes result in an ‘erroneous’ behaviour of the driver. Whether the driver assistance system can cope with such a circumstance

depends on to what extent the author-ity is given to the system. This paper discusses the need of ma-chine-initiated authority trading from the driver to the assistance system for assuring driver safety. A theoretical

framework is also given to describe and analyze the driver’s overtrust in and overreliance on such a driver as-sistance system.

TTTTOOOO WWWWHATHATHATHAT EEEEXTENTXTENTXTENTXTENT MMMMAYAYAYAY AAAASSISTANCESSISTANCESSISTANCESSISTANCE SSSSYSTEMSYSTEMSYSTEMSYSTEMS CCCCORRECTORRECTORRECTORRECT ANDANDANDAND PPPPREVENTREVENTREVENTREVENT

‘E‘E‘E‘ERRONEOUSRRONEOUSRRONEOUSRRONEOUS ’ B’ B’ B’ BEHAVIOUREHAVIOUREHAVIOUREHAVIOUR OFOFOFOF THETHETHETHE DDDDRIVERRIVERRIVERRIVER? ? ? ?

Toshiyuki Inagaki

University of Tsukuba, Department of Risk Engineering, Tsukuba 305-8573 Japan, [email protected]

MMMMANANANAN ----MACHINEMACHINEMACHINEMACHINE IIIINTEGRATIONNTEGRATIONNTEGRATIONNTEGRATION DDDDES IGNES IGNES IGNES IGN ANDANDANDAND AAAANALYS ISNALYS ISNALYS ISNALYS IS SSSSYSTEMYSTEMYSTEMYSTEM (MIDAS) (MIDAS) (MIDAS) (MIDAS) VVVV5: A5: A5: A5: AUGMENTATIONSUGMENTATIONSUGMENTATIONSUGMENTATIONS , , , ,

MMMMOTIVAT IONSOTIVAT IONSOTIVAT IONSOTIVAT IONS , , , , ANDANDANDAND DDDD IRECT IONSIRECT IONSIRECT IONSIRECT IONS FORFORFORFOR AAAAERONAUTICSERONAUTICSERONAUTICSERONAUTICS AAAAPPL ICATIONSPPL ICATIONSPPL ICATIONSPPL ICATIONS

Brian Gore NASA AMES RESEARCH CENTER Mail Stop 262-4

P.O. Box 1 Moffett Field, CA, USA 94035-0001, [email protected]

As automation and advanced tech-nologies are introduced into transport systems ranging from the Next Gen-eration Air Transportation System termed NextGen, to the advanced sur-face transportation systems as exem-plified by the Intelligent Transporta-tions Systems, to future systems de-signed for space exploration, there is an increased need to validly predict how the future systems will be vulner-able to error given the demands im-posed by the assistive technologies. One formalized approach to study the impact of assistive technologies on the human operator in a safe and non-

obtrusive manner is through the use of human performance models (HPMs). HPMs play an integral role when com-plex human-system designs are pro-posed, developed, and tested. One HPM tool termed the Man-machine Integration Design and Analysis System (MIDAS) is a NASA Ames Research Center HPM software tool that has been applied to predict human-system performance in various domains since 1986. MIDAS is a dy-namic, integrated HPM and simulation environment that facilitates the design, visualization, and computational evaluation of complex man-machine

system concepts in simulated opera-tional environments. The paper will discuss a range of aviation specific applications including an approach used to model human error for NASA’s Aviation Safety Program, and “what-if” analyses to evaluate flight deck technologies for NextGen operations. This chapter will culmi-nate by raising two challenges for the field of predictive HPMs for complex human-system designs that evaluate assistive technologies: that of (1) model transparency and (2) model validation.

The objective of the paper is to discuss the goals and scope of the EU project ISi-PADAS, the theo-retical backgrounds and assump-tions and the principal results achieved to date. The main idea of the project is to support the design and safety assessment of new gen-erations of assistance systems.

In particular, the development of autonomous actions is proposed, based on driver models able to pre-dict performances and reaction time, so as to anticipate potential incidental conditions. To achieve this objective two main integrated lines of development are proposed: (1) an improved risk based design

TTTTHEHEHEHE ISISISIS IIII ----PADAS PPADAS PPADAS PPADAS PROJECTROJECTROJECTROJECT ---- HHHHUMANUMANUMANUMAN MMMMODELLINGODELLINGODELLINGODELLING ANDANDANDAND SSSS IMULAT IONIMULAT IONIMULAT IONIMULAT ION TOTOTOTO SUPPORTSUPPORTSUPPORTSUPPORT HHHHUMANUMANUMANUMAN EEEERRORRRORRRORRROR

RRRR ISKISKISKISK AAAANALYS ISNALYS ISNALYS ISNALYS IS OFOFOFOF PPPPARTIALLYART IALLYART IALLYART IALLY AAAAUTONOMOUSUTONOMOUSUTONOMOUSUTONOMOUS DDDDRIVERRIVERRIVERRIVER AAAASS ISTANCESS ISTANCESS ISTANCESS ISTANCE SSSSYSTEMSYSTEMSYSTEMSYSTEMS

Pietro Carlo Cacciabue 1and Mark Vollrath,2

1 KITE Solutions, Via Labiena 93, 21014 Laveno Mombello, (Varese), Italia, [email protected]

2Technische Universität Braunschweig, Ingenieur- und Verkehrspsychologie, Gaußstr. 23, D-38106 Braunschweig. [email protected]

approach, able to account for a variety of human inadequate per-formances at different levels of cognition, and (2) the development of model for predicting correct and inadequate behaviour.


5


2Next Step Solutions, Rue Daussoigne-Mehul 20 4000 Liège, Belgium [email protected]

TTTTHEHEHEHE HUMAN PHUMAN PHUMAN PHUMAN PROJECTROJECTROJECTROJECT ––––

MMMMODELODELODELODEL ----BASEDBASEDBASEDBASED AAAANALYSISNALYSISNALYSISNALYSIS OFOFOFOF HHHHUMANUMANUMANUMAN EEEERRORSRRORSRRORSRRORS DURINGDURINGDURINGDURING AAAA IRCRAFTIRCRAFTIRCRAFTIRCRAFT CCCCOCKPITOCKP ITOCKP ITOCKP IT SSSSYSTEMYSTEMYSTEMYSTEM DDDDESIGNESIGNESIGNESIGN

1OFFIS Institute for Computer Science, Escherweg 2,26127 Oldenburg, GERMANY [email protected]

The objective of the HUMAN project is to develop a methodology with techniques and prototypical tools sup-porting the prediction of human errors in ways that are usable and practical for human-centred design of systems operating in complex cockpit environ-ments. The current approach of analysing systems is error prone as well as costly and time-consuming (based on engi-neering judgement, operational feed-back from similar aircraft, and simula-tor-based experiments). The HUMAN methodology allows to detect potential pilot errors more accurately and earlier(in the design) and with reduced effort.

The detection of errors is achieved by developing and validating a cognitive model of crew behaviour. Cognitive models are a means to make knowl-edge about characteristic human capa-bilities and limitations readily avail-able to designers in an executable form. They have the potential to auto-mate parts of the analysis of human errors because they offer the opportu-nity to simulate the interaction with cockpit systems under various condi-tions and to predict cognitive proc-esses like the assessment of situations and the resulting choice of actions including erroneous actions. In this way they can be used as a partial “substitute” for human pilots in early

development stages when design changes are still feasible and afford-able. Model- and simulation-based ap-proaches are already well-established for many aspects of the study, design and manufacture of a modern airliner (e.g., aerodynamics, aircraft systems, engines), for the very same objective of detecting potential problems earlier and reducing the amount of testing required at a later stage. HUMAN extends the modelling approach to the interaction of flight crews with cockpit systems.

Andreas Lüdtke 1, Denis Javaux 2 and the HUMAN Consortium

The objective of ITERATE is to develop and validate a unified model of driver behaviour (UMD) and driver interaction with innova-tive technologies in emergency situations. This model will be applicable to and validated for three surface transport modes: cars, ships, and trains. Drivers’ attitude, experience and culture are factors that will be considered together with influ-ences from the environment and the vehicle such as workload. The ITERATE model development has four main stages. The first stage has provided the

TTTTHEHEHEHE ITERATE ITERATE ITERATE ITERATE PROJECTPROJECTPROJECTPROJECT

OOOOVERVIEWVERVIEWVERVIEWVERVIEW , , , , THEORETICALTHEORETICALTHEORETICALTHEORETICAL FRAMEWORKFRAMEWORKFRAMEWORKFRAMEWORK ANDANDANDAND VAL IDAT IONVAL IDAT IONVAL IDAT IONVAL IDAT ION

overall model structure and the approach to ascertaining initial numeric values for the models as well as specifying the systems to be used in confirming and calibrat-ing the theoretical model. Stage 2 builds on this framework to carry out the experiments that have been specified for confirmation and cali-bration in a set of simulator stud-ies. In parallel, a software version of the UMD model will be built, and this version is intended for making predictions about how new systems will affect risk. In stage 3, when the initial set of experiments has been completed,

comprehensive analysis will be conducted to feed the model with the parameters produced from the experiments. Finally in stage 4, there will be a validation in which predictions from the software model will be compared with ob-servations of real drivers of cars and trains and navigators of ships. The present paper will discuss the ITERATE project general ap-proach, the theory behind the model and how it will be tested and validated.

Magnus Hjälmdahl, David Shinar and Oliver Carsten

VTI, Olaus Magnus väg 35, 581 95 Linköping, Sweden, [email protected], [email protected], [email protected]


6


AAAASSESSMENTSSESSMENTSSESSMENTSSESSMENT OFOFOFOF TRANSPORTATIONTRANSPORTATIONTRANSPORTATIONTRANSPORTATION SYSTEMSYSTEMSYSTEMSYSTEM RESIL IENCERESIL IENCERESIL IENCERESIL IENCE

Simon Enjalbert1,2,3, Frédéric Vanderhaegen1,2,3, Marianne Pichon1,2,3, Kiswendsida Abel Ouedraogo1,2,3 and Patrick Millot1,2,3

1Univ Lille Nord de France, F-59000 Lille, France 2UVHC, LAMIH, F-59313 Valenciennes, France

3CNRS, FRE 3304, F-59313 Valenciennes, France

{simon.enjalbert, frederic.vanderhaegen, marianne.pichon, kiswendsidaabel.ouedraogo, patrick.millot}@univ-valenciennes.fr

A transportation system like tram-way or train is a system in which the functions of the human and the machine are interrelated and neces-sary for the operation of the whole system according to Human-Machine System (HMS) definition. Both human and machines are sources of system reliability and causes of accident occurrences. Considering the human behaviour contribution to HMS resilience, resilience can only be diagnosed if the human actions improve the system performances and help to recover from instability.

inexperienced human operators. In railway transportation systems, traffic safety is the main perform-ance criterion to take into account. Based on this criterion, authors propose to evaluate an instantane-ous resilience indicator in order to assess the “local resilience” of HMS. As others performance crite-ria must be aggregated to reflect the whole studied HMS perform-ance, the “global resilience” of HMS will be defined.

Therefore, system resilience is the ability for a HMS to ensure performances and system stability whatever the context, i.e. after the occurrence of regular, unexpected or unprecedented disturbances. The COR&GEST platform is a railway simulation platform devel-oped in the LAMIH in Valen-ciennes which involves a miniature railway structure. In order to study the human behaviour during the train driving activities with or without any technical failure occurrences, an experimental protocol was built with several

There is a need to estimate impacts of proposed driver assistance sys-tems already at early stages of the system development process. Esti-mations of the impacts of new technologies have to be based on laboratory studies and modelling. This paper presents a traffic simu-lation based framework for estima-tion of the traffic system wide im-pacts of driver assistance systems.

EEEEST IMATINGST IMATINGST IMATINGST IMATING TRAFF ICTRAFF ICTRAFF ICTRAFF IC SYSTEMSYSTEMSYSTEMSYSTEM WIDEWIDEWIDEWIDE IMPACTSIMPACTSIMPACTSIMPACTS OFOFOFOF DRIVERDRIVERDRIVERDRIVER ASS ISTANCEASS ISTANCEASS ISTANCEASS ISTANCE SYSTEMSSYSTEMSSYSTEMSSYSTEMS USINGUS INGUS INGUS ING

TRAFF ICTRAFF ICTRAFF ICTRAFF IC SIMULAT IONSIMULAT IONSIMULAT IONSIMULAT ION

The framework includes a two step methodology. In the first step of the analysis, the considered driver assistance system’s impact on driver behaviour is observed. The second step of the analysis consist of traffic simulation modelling taking into account the system functionality as well as the ob-served driver behaviour of the con-sidered driver assistance system.

Driver behaviour studies for use of the data for traffic simulation mod-eling is discussed and traffic simu-lation modelling of different types of driver assistance systems is ex-emplified by modelling of an over-taking assistant, of in-vehicle vir-tual rumble strips and of adaptive cruise control.

Andreas Tapani

Swedish National Road and Transport Research Institute (VTI), SE-581 95 Linköping, Sweden

[email protected]

3. ABSTRACTS OF ITERATE PAPERS3. ABSTRACTS OF ITERATE PAPERS3. ABSTRACTS OF ITERATE PAPERS3. ABSTRACTS OF ITERATE PAPERS


The purpose of this paper is to dis-cuss changes in the train driver’s work task, in terms of prospects and problems, concerning how to make good use of the new techno-logical systems which at present are being introduced across Europe

HHHHUMANUMANUMANUMAN FFFFACTORSACTORSACTORSACTORS EEEENGINEERINGNGINEERINGNGINEERINGNGINEERING ININININ TRAINTRAINTRAINTRAIN CABCABCABCAB DES IGNDES IGNDES IGNDES IGN –––– PPPPROSPECTSROSPECTSROSPECTSROSPECTS ANDANDANDAND PROBLEMSPROBLEMSPROBLEMSPROBLEMS

but also to manage new risk situa-tions involving user interaction and organizational issues. Examples of new risk situations for the driver involve information overload and divided attention. In conclusion, this paper highlights the need for a

systematic work process within the railway sector in general to man-age MTO or human factors issues. The concept of Human Factors Engineering (HFE) is also used synonymously to describe such systematic processes.

Lena Kecklund, Aino Mowitz and Markus Dimgard

MTO Säkerhet Hornsbruksgatan 28,Hornsbruksgatan 28,117 34 Stockholm, SWEDEN, [email protected], [email protected], [email protected]

7


Driver behaviour can be modelled in one of two approaches: 'Descriptive' models that describe the driving task in terms of what the driver does, and 'Functional' models that attempt to explain why the driver behaves the way he/she does, and how to predict drivers' performance in demanding and routine situations. Demanding situations elicit peak performance capabilities, and routine situations elicit typical (not necessarily best) behaviour. It seems that the opti-mal approach might be a hybrid of several types of models, extracting the most useful features of each. In recent years, a variety of driver support and information manage-ment systems have been designed and implemented with the objec-tive of improving safety as well as performance of vehicles. To predict the impact of various

assistance systems on driver behaviour predictive models of the interaction of the driver with the vehicle and the environment are necessary. The first step of the ITERATE project is to critically review exist-ing Driver-Vehicle-Environment (DVE) models and identify the most relevant drivers' parameters and variables that need to be in-cluded in such models: (a) in dif-ferent surface transport modes (this paper deals with road vehicles only, other transport domains are detailed in D1.1 & D1.2 of the IT-ERATE project), and (b) in differ-ent safety critical situations. On the basis of this review, we propose here a Unified Model of Driver behaviour (UMD), that is a hybrid model of the two approaches. The model allows for individual differences on pre-specified

dimensions and includes the vehi-cle and environmental parameters. Within the ITERATE project this model will be used to support safety assessment of innovative technologies (based on the abili-ties, needs, driving style and ca-pacity of the individual drivers). In this brief paper we describe only the behaviour of a single test driver, while the environment and vehicle are defined as parameters with fixed values (and detailed in D1.2 of the ITERATE project). The selected driver characteristics (and variables used to measure them) are culture (Country), atti-tudes/personality (Sensation Seek-ing), experience (Hazard Percep-tion Skills) , driver state (Fatigue), and task demand (Subjective work-load).

RRRREVIEWEVIEWEVIEWEVIEW OFOFOFOF MMMMODELSODELSODELSODELS OFOFOFOF DDDDRIVERRIVERRIVERRIVER BEHAVIOURBEHAVIOURBEHAVIOURBEHAVIOUR ANDANDANDAND DEVELOPMENTDEVELOPMENTDEVELOPMENTDEVELOPMENT OFOFOFOF AAAA UUUUNIF IEDNIF IEDNIF IEDNIF IED DDDDRIVERRIVERRIVERRIVER BBBBEHAVIOUREHAVIOUREHAVIOUREHAVIOUR

MODELMODELMODELMODEL FORFORFORFOR DRIV INGDRIV INGDRIV INGDRIV ING ININININ SAFETYSAFETYSAFETYSAFETY CRITICALCRITICALCRITICALCRITICAL SITUAT IONSSITUAT IONSSITUAT IONSSITUAT IONS ....

David Shinar and Ilit Oppenheim

Ben-Gurion University of the Negev, Beer Sheva, 84105, Israel [email protected], [email protected]

FFFFROMROMROMROM THEORETICALTHEORETICALTHEORETICALTHEORETICAL MODELMODELMODELMODEL TOTOTOTO EXPERIMENTALEXPERIMENTALEXPERIMENTALEXPERIMENTAL DATADATADATADATA : A : A : A : A STRUCTUREDSTRUCTUREDSTRUCTUREDSTRUCTURED APPROACHAPPROACHAPPROACHAPPROACH TOTOTOTO DES IGNDES IGNDES IGNDES IGN

EXPERIMENTSEXPERIMENTSEXPERIMENTSEXPERIMENTS TOTOTOTO SEEDSEEDSEEDSEED AAAA MODELMODELMODELMODEL OFOFOFOF VEHICLEVEHICLEVEHICLEVEHICLE OPERATIONOPERATIONOPERATIONOPERATION WITHWITHWITHWITH NEWNEWNEWNEW SYSTEMSSYSTEMSSYSTEMSSYSTEMS

Yvonne Barnard, Oliver Carsten and Frank Lai Institute for Transport Studies

University of Leeds LEEDS, UK, LS2 9JT [email protected], [email protected], [email protected]

In this paper we will discuss a methodology developed and ap-plied in the European ITERATE project with the objective of designing experiments that will provide data to seed the ITERATE theoretical model of operator behaviour in different surface transport modes: road vehicles, rail transport and ships.

A structured approach was taken involving seven steps: (1) Selec-tion of operator support systems to be studied; (2) formulation of hy-potheses on the effects of the op-erator parameters from the model on the interaction with the sys-tems; (3) final system selection; (4) operationalisation of operator parameters and identification of

ways to measure them; (5) devel-opment of scenarios; (6) develop-ment of experimental set-ups; and (7) specification of simulators and experiments .


8


4. SIMULATOR4. SIMULATOR4. SIMULATOR4. SIMULATOR

The main objective of ITERATE project is to develop and validate a unified model of driver behaviour (UMD) and driver interaction with innovative technologies in emer-gency situations. This model must be applicable and validated for all surface transport modes, starting on the reasonable assumption that the underlying factors influencing human behaviour are constant. In this context, it is clear the fun-damental role of specific empirical experiments. At this purpose, an important step of work in the pro-ject is the performance of simula-tor experiments, which allow to test a number of relevant cases and scenarios. These experiments are carried out for cars and trains and some of the best simulators in Europe are used to make sure that the results are accurate and valid. In particular, a car and train simu-

Figure 4. Driver Simulator for Car

simulator is running on a HP Z400 workstation computer with an ATI 5870 Graphics Card. The images are displayed on a 40 inch screen Samsung Sync-Master 400MX-2. The dri1er DMI is displayed on a ViewSonic VA1616W screen. A GameRacer seat is provided to the driver. When driving in car-mode, a Logi-tec G27 steering wheel and pedals are used. When driving in train-mode use RailDriver controller is used. In the experiments some detailed behavioural parameters, such as handling of controls/steering whe-el, lane position and braking/stopping behaviour, are analyzed The simulator is currently being shipped between partners thus allo-wing the representation of different cultural regions. In the next steps of work, the re-sults from simulator experiments will be used to quantify the pa-rameters needed for the ITERATE model and to establish the links between the parameters and the driver characteristics.

Figure 3. Driver Simulator for Train

lator has been developed by VTI and by the University of Leeds and all participants to the HMAT-2010 Workshop have been invited to visit it. From a technical point of view, the

9

5. TABLE OF PUBLICATIONS5. TABLE OF PUBLICATIONS5. TABLE OF PUBLICATIONS5. TABLE OF PUBLICATIONS

Title Authors Type

A simple simulation predicting driver behaviour, attitudes and errors

A. Amantini P. Cacciabue

Paper presented at the HCI International 2009 and included in the conference proceedings published by Springer

From theoretical model to experimental data: A structured approach to design experiments to seed a model of vehicle operation with new systems

Y. Barnard O. Carsten F. Lai

HMAT 2010 - Paper

Estimating traffic system wide impacts of driver assistance systems using traffic simulation

A. Tapani HMAT 2010 - Paper

Review of Models of Driver behaviour and development of a Unified Driver Behaviour model for driving in safety critical situations

D. Shinar I. Oppenheim

HMAT 2010 - Paper

Human Factors Engineering (MTO) in train drivers cab design – Prospects and problems

L. Kecklund A. Mowitz M. Dimgard

HMAT 2010 - Paper

Assessment of transportation system resilience

S. Enjalbert F. Vanderhaegen M. Pichon K. Ouedraogo P. Millot

HMAT 2010 - Paper

The ITERATE Project - Overview, theoretical framework and validation

M. Hjälmdahl D. Shinar O. Carsten B. Peters

HMAT 2010 - Keynote

Automated driving, secondary task performance and situation awareness

N. Merat F. Lai H. Jamson O.Carsten

Paper presented at the Human Factors and Ergonom-ics Society Europe Chapter Annual Conference, 14-16 October 2009, Linköping, Sweden.

Preparing field operational tests for driver support systems: a research oriented approach

O. Carsten Y. Barnard


Spotting sheep in Yorkshire: using eye-tracking for studying situation awareness in a driving simulator

Y. Barnard F. Lai


Semi-automated driving: how does the supported task affect driver response?

O.Carsten F. Lai Y. Barnard

Paper presented at the IFAC HMS 2010 Symposium, Valenciennes, France.

A simple model of driver behaviour to sustain design and safety assessment of automated systems in automotive environments

P.C. Cacciabue O. Carsten

Applied Ergonomics 41, pp. 187–197.

A numerical tool for reproducing driver behaviour: Experiments and predictive simulations

M. Casucci M. Marchitto P.C. Cacciabue

Applied Ergonomics 41, pp. 198–210.

10

RRRREADEADEADEAD OUROUROUROUR NEWSLETTERNEWSLETTERNEWSLETTERNEWSLETTER

Deta i led ar t ic les wi l l in form you about

Recent news, incoming events, achieved results


Project CoordinatorProject CoordinatorProject CoordinatorProject Coordinator Magnus Hjälmdahl - VTI [email protected]

DisseminationDisseminationDisseminationDissemination Aladino Amantini - KITE Solutions

[email protected] Giuseppe Russo - KITE Solutions

[email protected]

Contacts

VVVV IS ITIS ITIS ITIS IT OUROUROUROUR WEBSITEWEBSITEWEBSITEWEBSITE : : : : http://www.iteratehttp://www.iteratehttp://www.iteratehttp://www.iterate----project.euproject.euproject.euproject.eu Here you can f ind background mater ia l concern ing

Project st ructure, partners, aims and targets

Consortium

IT for Error Remediation And Trapping Emergencies ... · PDF fileModelling in Assisted...

Documents

Transcript of IT for Error Remediation And Trapping Emergencies ... · PDF fileModelling in Assisted...