Guide to Applying Human Factors Methods: Human Error and Accident Management in Safety-Critical Systems

353 Pages·2004·11.92 MB·English

by  Pietro Carlo Cacciabue PhD (auth.)

Guide to Applying Human Factors Methods
Pietro Carlo Cacciabue
Guide to Applying
Human Factors
Methods
Human Error and Accident Management in
Safety-Critical Systems
~ Springer
Pietro Carlo Cacciabue,PhD
EC,Joint Research Centre
Institutefor the Protection and Securityof the Citizen
Varese,1-21020
Italy
BritishLibrary CataloguinginPublicationData
Cacciabue,Pietro C.
Guidetoapplyinghumanfactors methods:humanerrorand accident management
insafety-criticalsystems
I. Human engineering
I. Title
620.8'2
LibraryofCongress Cataloging-in-PublicationData
Cacciabue,Pietro C.
Guidetoapplyinghumanfactors methods:Pietro CarloCacciabue.
p. em.
Includes bibliographicalreferences and index.
I. Humanengineering. I. Title.
TA166.C332004
620.8'2- dc22
2004044972
ISBN 978-1-84996-898-0 ISBN 978-1-4471-3812-9 (eBook)
DOI 10.1007/978-1-4471-3812-9
Apart fromanyfairdealingforthepurposesofresearchorprivatestudy,orcriticismorreview,asper
mitted underthe Copyright,Designs and PatentsAct1988,this publication mayonly bereproduced,
storedor transmitted,inanyform orbyanymeans,with the priorpermission inwritingofthe pub
lishers,orinthecaseofreprographicreproduction inaccordancewiththeterms oflicencesissuedby
theCopyrightLicensingAgency.Enquiriesconcerningreproductionoutsidethosetermsshouldbesent
tothe publisher.
springeronline.com
©Springer-VerlagLondon2004
OriginallypublishedbySpringer-VerlagLondonBerlinHeidelbergin2004
Softcoverreprintofthehardcover1stedition2004
Theuseofregisterednames,trademarksetc.inthispublicationdoes notimply,evenintheabsenceof
aspecificstatement,that such names areexempt fromthe relevantlawsand regulationsand therefore
freeforgeneral use.
Thepublishermakes no representation,express or implied,with regard tothe accuracyofthe infor
mationcontainedinthis book and cannotaccept anylegalresponsibilityorliability foranyerrorsor
omissionsthat maybemade.
Typesetting:SNPBest-setTypesetterLtd.,HongKong
34/3830-543210Printedonacid-freepaperSPIN10896211
To Cathy
Contents
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
1 Introduction.............................................. 1
2 Elements of Human-Machine Systems ......................... 9
3 Theories and Methods for Human-Machine Systems. . . . . . . . . . . . . . 48
4 Guidelines for the Application of HF Methods in HMS . . . . . . . . . . . . . 105
5 Application of HERMES for the Design of HMI Systems:
Anticollision Warning in Automotive Environments. . . . . . . . . . . . . . . 147
6 Application of HERMES in Training: A Human Factors
Course in the Aviation Domain ............................... 190
7 Application of HERMES for Safety Assessment: The Safety
Audit of a Large Railway Organisation . . . . . . . . . . . . . . . . . . . . . . . . . . 230
8 Application of HERMES for Accident Investigation:
The Case of a Thermoelectric Power Plant ...................... 293
Bibliography ................................................. 332
Index.... ...... ... ........ ..... ...... ......... ...... ...... .. 342
vii
Acknowledgments
This book contains work carried out in collaboration with a group of outstanding
people over aperiod of 10 years.
What makes the Human Factors Group special is that it was a set of constantly
changing temporary researchers, scientific visitors, and PhD or graduating stu
dents, who remained for a number of years to work in the group or collaborated
from their universities and labs before going on either to start a family or to
become researchers, managers, university professors, engineers, or psychologists.
All these people shared one common characteristic: They fell in love with the study
ofhuman factors. The professor who guided me in my graduation thesis in nuclear
engineering was very careful in warning me that neutron physics is a science that
traps people and makes them want to do nothing else in their professionallife. This
is what happened to me. However, all that changed when I discovered human
factors in the early 1980s. Nowadays, I say to all the students I work with: "Be
careful, because you may fall in love with human factors."
This common interest and locus of attention is what makes this group of people
remarkable - probably not unique, but certainly exceptional. The results of the
work that we performed together are condensed in this book.
It is impossible to name all the members of this ever-changing group without
risking that someone may be left out. It is much better to thank the Human Factors
Group as a whole.
I am in debt to all of them for their support, help, contribution, enthusiasm, and
encouragement, especially when the group has had to be temporarily dismantled
at various stages. But the group has always been rebuilt with new faces, brains, and
hearts because the work may have sometimes been interrupted, but the spirit
never died.
Thank you, Human Factors Group - this book belongs to us all.
Pietro Carlo Cacciabue
Varese, Italy
ix
Introduction
It could be argued that Human Factors (HF) is not a pure science, as it is trans
versal to other well-established sciences, such as physics, mathematics, psychology,
and sociology.
Stricdy speaking, this is correct as Human Factors is not self-referential, as all pure
sciences are, and does not base its theories on unique and exclusive conservation
principles and postulates.
Like all modern sciences, such as computing, economics, communication, and
management, Human Factors spans several existing theoretical backgrounds and
requires strong connections with practical areas of application. This ability repre
sents an essential attribute for ensuring efficiency and effectiveness in real working
environments. In particular, Human Factors extend over four essential domains,
namely, engineering, psychology, sociology, and computer science. Each of these
domains requires a profound theoretical knowledge and is equally linked to prac
tical applications in technology and everyday life.
At the same time, it is limiting to identify Human Factors as a discipline, as it goes
far beyond the mere application of engineering, psychology, sociology, and com
puter science principles, and requires blending existing theoretical methods in
all these four fields, generating new and specific theoretical formulations and
paradigms. In this way, it becomes possible to represent working contexts and
sociotechnical aspects in a theoretical form, which then needs further simplifica
tion and elaboration in order to develop practical applications and quantifications
for implementation in real systems and for the assessment of real working
contexts.
Human factors, together with computer science, is the most relevant science that
has been developed over the past 50 years, as a consequence of the technological
development and the role assigned to human beings in the management of
pro ces ses that have progressively become more and more complex and demand
ing in terms of control ability and skill. This latter aspect has increased attention
on humans as the "weak point" in accident chains and in process control, and has
generated the need to study more accurately the interaction of humans with
machines.
In addition to these aspects, Human Factors bears a fundamental connection with
the "mother" of all sciences, i.e., philosophy. This connection is quite obvious as
P. C. Cacciabue, Guide to Applying Human Factors Methods
© Springer-Verlag London 2004
2 Guide to Applying Human Factors Methods
the core subjects of Human Factors are human beings and their behaviour, which
depends on their thinking and beliefs. This is indeed the subject matter of philos
ophy and consequently, correlates with the major philosophers and thinkers of the
past and more re cent times.
1.1 Aim of This Book
This book is dedicated to readers that have already acquired basic knowledge in
the domain of human-machine interaction and cognitive ergonomics. It does not
offe r an introductory overview and basic understanding in Human Factors, as the
literature is rich with very valuable textbooks that offer preeisely this basic knowl
edge (Sheridan and Ferrell, 1974; Sanders and McCormick, 1976; Rouse, 1980;
Kantowitz and Sorkin, 1983; Rasmussen, 1986).
The aim is to guide readers how to perform Human Factors analysis and how
to select and implement existing methods and techniques for solving practical
problems.
The process of selection and implementation is usually assoeiated with a method
ology that offers a roadmap for practitioners and requires formal and accurate
application in order to ensure that models, methods, and data are adequately
selected and fruitfully exploited. The results of applying such a methodology are
consistently assoeiated with implementations, recommendations, and assessments
aimed at improving and augmenting systems safety.
This is why the book rotates around a reference methodology, called Human Error
Risk Management in Engineering Systems (HERMES) that is described in detail in
the first part and guides the practical applications to real case studies in the second
part.
Human factors methods are applied in four main areas of application, namely,
design, safety assessment, training, and aceident investigation. The main domains
of application of Human Factors are energy production (nuclear and conven
tional), transportation systems (aviation, railway, automotive, maritime), medi
eine, economic system, chemical and petrochemical environments, manufacturing,
and economical systems.
These domains are usually based on large complicated plants and organisations.
Consequently, the scenarios of practical applications of a Human Factors analysis
are very complex and demand relatively extensive pro ces ses and lengthy imple
mentations of methods and techniques. A crueial requirement that derives from
such complexity is that the methodology guiding the Human Factors application
must be strictly and formally put into practice, in order to avoid trivial imple
mentations leading to unrealistic or even misleading resuIts.
1. Introduction 3
1.2 How to Read This Book
This book is divided in two parts, dedicated respectively to the theoretical descrip
tion of the methodology and associated methods and techniques, and to the devel
opment of a number of relevant real cases in various areas of applications and in
different domains.
In particular, in the first part of the book, i.e., Chapters 2, 3, and 4, the theoretical
and methodological bases of Human Factors are discussed with reference to
design, safety assessment, training, and accident investigation. The basic hypoth
esis and theoretical stands that sustain the HERMES methodological framework
are presented. The most common methods and models for describing and study
ing human-machine systems are briefly discussed and compared. The procedures
for implementing HERMES in different areas of application are described.
In the second part, i.e., Chapters 5, 6, 7, and 8, four complex cases in real domains
are described in detail, covering aH steps of implementation and showing how the
HERMES methodology has been applied.
The two parts of the book can in principle be read separately. The expert Human
Factors reader, interested primarily in practical implementation, can selectively
read chapters of the second part and may refer back to specific sections in the first
part that describe the associated theoretical and methodological background.
The less expert readers may initiaHy pay attention to the first part of the book,
where theories, paradigms, and basic principles of Human Factors are presented.
In this case, the detailed application and complexity of the case studies are of minor
interest, even if reading these chapters offers a feeling of the amount of work that
practical applications require.
Hopefully, the basic inference that can be drawn from reading the whole book, or
even from considering either parts separately, is that the application of Human
Factors to practical problems demands an accurate work of analysis and the
implementation of formal methods and techniques, which are weH established and
consolidated. Therefore, Human Factors problems cannot be solved by simply
applying common sense and practical experience.
Moreover, the intrinsic dynamic and evolutionary nature of organisations, with the
turnover of personnel and management, and the chan ging of technical and control
systems as weH as production, requires that the Human Factors issues are to be
dealt with not only once in the life of a system, but demand constant awareness
and recurrent application of audits. This grants that adequate safety levels are
reached and maintained at all times.
4 6uide to Applying Human Factors Methods
1.3 Brief Summary of Each Chapter
Chapter2
Chapter 2 contains the basie definitions and all theoretieal foundations of the
methodology.
In partieular, the basie definitions of Human Factors (HF), Human-Machine
Systems (HMS), and Human Errors (HE) are worked out in relation to the vast lit
erature available that has been developed over many years. A number of related
conceptual subjects are elaborated from these definitions. In partieular, the
sociotechnical elements that exist in any working context are defined.
The chapter then concentrates on human error and human error management and
these concepts are expanded to embrace the vital issue of accident management.
This is the critieal safety concern of all modern technological systems.
The concept of Human Error and Accident Management (HEAM) is developed. In
partieular, five "standpoints" are defined in association with the HEAM definition.
These standpoints represent the premises and issues that have to be resolved and
established clearly in any HF analysis, in order to restriet the scope of the outcome
to well-defined objectives and goals, and to support the HF analysts both during
the selection of methods and models and during their implementation for specific
studies.
The HEAM concept and the associated standpoints lie at the base of the HERMES
methodology, whieh is the core subject of this book and is described as a direct
implementation of the HEAM measures and requirements for safety studies in real
working environment and organisations. The relation of the HERMES metho
dology with the modem concept of ecology and cognitive ergonomies is also
discussed.
Chapter 3
Chapter 3 contains a short review of theories and methods for studying Human
Machine Interaction (HMI).
Firstly, models and taxonomies that represent human behaviour are described.
These are the reference paradigms that can be found in literature and permit rep
resentation of the vast majority of working contexts and conditions. However,
many other models exist, and a quite extensive literature about them is reported
in the references of the book.
In addition to models and taxonomies of human behaviour, the other basie
methods supporting HEAM analysis and the HERMES methodology are consid
ered in this chapter, namely, ethnographie studies, cognitive task analysis, root
cause analysis, quantitative risk assessment, and recurrent safety audit. The Ethno
graphie Studies (ES) are the set of theoretieal and empirieal methods for evaluat
ing real working contexts. ES are discussed with reference to retrospective and
prospective types of analysis.