Table of Contents

Introduction ……………………………………………………………. 1
1.1 Aviation as a complex, hierarchical system …………………………………………………….. 3
1.2 The general structure of the CATS model ……………………………………………………….. 5
1.3 Safety management influences on risk ……………………………………………………………. 8
1.4 Scope, approaches and outline of this thesis ………………………………………………….. 10
2 Safety management model …………………………………….. 13
2.1 Dutch safety management models ………………………………………………………………… 14
2.1.1 I-Risk ……………………………………………………………………………………………… 14
2.1.2 ARAMIS ………………………………………………………………………………………… 17
2.1.3 WORM …………………………………………………………………………………………… 23
2.1.4 CATS ……………………………………………………………………………………………… 24
2.1.5 Internal conclusions for this section ……………………………………………………. 26
2.2 Mapping the Dutch model with other aviation management models …………………. 27
2.2.1 Human Factors Analysis and Classification System (HFACS) ………………. 28
2.2.2 SoTeRiA …………………………………………………………………………………………. 33
2.3 Safety management system in practice ………………………………………………………….. 35
2.3.1 International SMS standards ……………………………………………………………… 35
2.3.2 Interviews with airlines …………………………………………………………………….. 38
2.4 Findings and suggestions …………………………………………………………………………….. 42
3 Human performance……………………………………………… 45
3.1 Cognitive frameworks of task performance and human error …………………………… 45
3.1.1 Underlying mechanism …………………………………………………………………….. 45
3.1.2 The factors essential for linking to safety management …………………………. 46
3.2 Accident/incident investigation schemes ……………………………………………………….. 51
3.2.1 Human Factors Analysis and Classification System (HFACS) ………………. 52
3.2.2 ICAO Accident/Incident Data Reporting System (ADREP /ECCAIRS) …. 53
3.2.3 Line Operations Safety Audit (LOSA) ………………………………………………… 56
3.2.4 Interim conclusion ……………………………………………………………………………. 57
3.3 Human factors in reliability-oriented techniques or PSA driven methods ………….. 60
3.3.1 Causal Model for Air Transport Safety (CATS) …………………………………… 60
3.3.2 Socio-technical risk analysis (SoTeRiA) …………………………………………….. 63
3.3.3 Eurocontrol Integrated Risk Picture (IRP) …………………………………………… 65
3.3.4 Conclusion on PSA models ……………………………………………………………….. 65
3.4 Overall Conclusions ……………………………………………………………………………………. 65
4 Technical performance ………………………………………….. 67
4.1 Design and manufacturing …………………………………………………………………………… 69
4.2 Flight crew operation (Malfunction due to crew action or inaction) ………………….. 71
4.3 Maintenance and inspection ………………………………………………………………………… 73
4.4 Proposals …………………………………………………………………………………………………… 79
5 Quantification methods of the SMS ……………………….. 81
5.1 Requirements and challenges to integrate managerial influences in quantified risk
modelling ………………………………………………………………………………………………….. 81
5.2 How much data is available? ……………………………………………………………………….. 82
5.2.1 ADREP …………………………………………………………………………………………… 83
5.2.2 LOSA …………………………………………………………………………………………….. 87
5.2.3 EU-OPS and IOSA …………………………………………………………………………… 90
ii
5.3 System Dynamics ………………………………………………………………………………………. 94
5.4 Bayesian Belief Nets (BBNs) ………………………………………………………………………. 96
5.4.1 General description ………………………………………………………………………….. 96
5.4.2 Discrete BBNs …………………………………………………………………………………. 98
5.4.3 Distribution free continuous BBNs …………………………………………………… 100
5.4.3.1 Lesson learned from CATS …………………………………………………….. 101
5.5 An additional method for quantification ………………………………………………………. 105
5.5.1 Paired comparisons ………………………………………………………………………… 106
5.5.2 Paired comparisons combined with distribution free continuous BBNs
approach ……………………………………………………………………………………….. 108
5.6 Summary and conclusion …………………………………………………………………………… 113
6 Using paired comparisons to quantify the effect of
management influences in CATS …………………………. 115
6.1 First experiment ……………………………………………………………………………………….. 115
6.1.1 Experiment design ………………………………………………………………………….. 115
6.1.2 Elicitation procedure ………………………………………………………………………. 118
6.1.3 Data collection and analyses ……………………………………………………………. 119
6.1.4 Management effect on the risk …………………………………………………………. 124
6.1.5 Discussion …………………………………………………………………………………….. 125
6.2 Second experiment ……………………………………………………………………………………. 126
6.2.1 Experiment design ………………………………………………………………………….. 127
6.2.2 Elicitation procedure ………………………………………………………………………. 129
6.2.3 Data collection and analyses ……………………………………………………………. 130
6.2.4 Findings and discussions …………………………………………………………………. 132
6.3 Findings and suggestions …………………………………………………………………………… 133
7 Integrated model and proposal for the future……….. 135
7.1 A generic hierarchical control model for aviation safety ……………………………….. 137
7.1.1 The treatment of hierarchical relations: A general structured approach …. 137
7.1.2 A detailed modelling of system levels: insights from previous chapters … 138
7.2 New version of the Dutch SMS ………………………………………………………………….. 142
7.2.1 A generic structure per delivery system: A simplified model ……………….. 142
7.2.2 New and clarified delivery systems ………………………………………………….. 144
7.2.2.1 Competence and suitability ……………………………………………………… 145
7.2.2.2 Suitability ……………………………………………………………………………… 147
7.2.2.3 Manpower planning and availability ………………………………………… 149
7.2.2.4 Workload………………………………………………………………………………. 150
7.2.2.5 Procedures, rules, checklists, and goals …………………………………….. 152
7.2.2.6 Communication and coordination …………………………………………….. 154
7.2.2.7 Man-machine interface …………………………………………………………… 156
7.2.2.8 Commitment to safety …………………………………………………………….. 158
7.3 Enhancing PSF quantification ……………………………………………………………………. 166
7.3.1 Quantifying qualitative variables ……………………………………………………… 167
7.3.2 Quantify distributions for qualitative variables …………………………………… 167
7.4 How to improve the availability of the data in aviation? ………………………………… 168
7.5 Summary …………………………………………………………………………………………………. 169
8 Summary, conclusions and recommendations ………. 171
8.1 State of the art in the Dutch management model and need for additional
improvement ……………………………………………………………………………………………. 171
iii
8.1.1 A review of the history of the development of the Dutch model and issues
remaining to be solved ……………………………………………………………………. 172
8.1.2 Comparative validation of the Dutch model ………………………………………. 173
8.2 The underlying causes that contribute to human performance and aircraft
deficiencies in aviation and need for additional improvement in modelling ……… 174
8.2.1 Human performance model ……………………………………………………………… 174
8.2.2 Technical performance model ………………………………………………………….. 175
8.3 Available data and their problems in risk modelling in aviation ……………………… 178
8.4 Quantification methods of SMSs and their problems in risk modelling in aviation
………………………………………………………………………………………………………………. 178
8.4.1 System Dynamics …………………………………………………………………………… 178
8.4.2 Discrete BBNs ……………………………………………………………………………….. 179
8.4.3 Distribution free continuous BBNs …………………………………………………… 179
8.4.4 Supplementary method (combining paired comparisons with distribution
free continuous BBNs) ……………………………………………………………………. 180
8.5 Improving safety management modelling and its quantification in aviation …….. 181
8.5.1 Clarify the hierarchical relations between the SMS and operations ………. 181
8.5.2 Improve the detailed modelling of the system levels …………………………… 182
8.5.3 Clarify delivery systems into a generic structure ………………………………… 183
8.6 Final conclusion ……………………………………………………………………………………….. 184
8.7 Limitations ………………………………………………………………………………………………. 184
8.8 Recommendations for future work ……………………………………………………………… 185
8.8.1 For qualitative modelling ………………………………………………………………… 185
8.8.2 For the methodology of quantification ………………………………………………. 186
References ………………………………………………………………… 189
Appendices ……………………………………………………………….. 197
Appendix A: Definition of delivery systems of ARAMIS ……………………………………. 198
Appendix B: Selected example of unsafe acts and preconditions in HFACS ………….. 202
Appendix C: Human resource functions (Ostroff, 1995) ……………………………………… 203
Appendix D: LOSA “error” codes and “threat” codes …………………………………………. 206
Appendix E: ADREP and delivery systems mapping ………………………………………….. 214
Appendix F: Influence diagram calculations (Embrey, 1992) ……………………………….. 222
Appendix G: Management actions for “fatigue”, “weather”, and “workload” ………… 223
Appendix H: Distribution of rainfall rate and Distribution of number of time the crew
members have to refer to the A/E procedure ………………………………………………… 225
Summary ………………………………………………………………….. 227
Samenvatting ……………………………………………………………. 233
Acknowledgements ……………………………………………………. 239
About the author ………………………………………………………. 241
NGInfra PhD thesis series on infrastructures …………….. 243

Abstract

Aviation accidents result from a combination of many different causal factors ( human errors, technical failures, environmental and organisational influences). Increasing interest over the past two decades in causal modelling of organisational factors has been motivated by the desire to understand these fundamental causes and their influences in risk. Although there is a need for system-wide accident models in air transport, such models are currently lacking. Causal Modelling of Air Transportation System (CATS) was one of the first projects to develop such a model in aviation. Based on the experience of CATS, this PhD research examines the place and role of the human and management models and their quantification in aviation. This study reveals several challenges in respect to safety management modelling, including availability of data, and techniques for the management quantification. New insights found in this research were taken on board to develop a generic hierarchical control model for aviation safety and a list of human and technical factors to be treated in risk modelling in aviation. A new way of quantifying safety management in risk model is also proposed. Several recommendations were made for an extension of risk modelling in CATS, or in the other research with similar research objectives.

Roadmap onderwerpen

Veerkrachtig ontwerp en onderhoud

Terug naar overzicht