A Formal Framework for the Analysis of Human-Machine Interactions
Première édition
There are more and more automated systems with which people are led to interact everyday. Their complexity increases, and badly designed systems may result in automation surprises. The contribution of this thesis is a formal analysis framework to assess whether a system is prone to potential automation surprises in an interaction. Lire la suite
There are more and more automated systems with which people are led to interact everyday. Their increasing complexity makes it harder for operators to drive them safely, in particular because badly designed systems may result in automation surprises. Several accidents are due to such surprising situations, as testified by real accidents. Examples include the Three Mile Island nuclear meltdown, the lethal radiation doses administered by the Therac 25 medical device, or the shot down of the aircraft of the KAL007 flight.
The contribution of this thesis is a formal analysis framework to assess whether a system is prone to potential automation surprises in an interaction. Potential automation surprises are captured by the full-control property. The minimal fullcontrol conceptual model generation problem consists in finding a minimal conceptual model that allows full-control of the system. The existence of such conceptual model is characterised by the fc-deterministic property. The generated models can be used to generate artifacts, such as user manuals. Three algorithms are proposed: the
first one uses three-valued deterministic finite automata that characterise the fullcontrol property in terms of traces; the second one uses a variant of the Paige-Tarjan reduction algorithm, and the third one uses the L* active learning algorithm. The proposed framework has been tested on various examples, among which a large case study of an autopilot coming from ADEPT, a toolset to support designers in the early design phases of automation interfaces. Experiences show that the method proposed in this thesis can identify existing automation surprises and also find new ones.
There are more and more automated systems with which people are led to interact everyday. Their increasing complexity makes it harder for operators to drive them safely, in particular because badly designed systems may result in automation surprises. Several accidents are due to such surprising situations, as testified by real accidents. Examples include the Three Mile Island nuclear meltdown, the lethal radiation doses administered by the Therac 25 medical device, or the shot down of the aircraft of the KAL007 flight.
The contribution of this thesis is a formal analysis framework to assess whether a system is prone to potential automation surprises in an interaction. Potential automation surprises are captured by the full-control property. The minimal fullcontrol conceptual model generation problem consists in finding a minimal conceptual model that allows full-control of the system. The existence of such conceptual model is characterised by the fc-deterministic property. The generated models can be used to generate artifacts, such as user manuals. Three algorithms are proposed: the first one uses three-valued deterministic finite automata that characterise the fullcontrol property in terms of traces; the second one uses a variant of the Paige-Tarjan reduction algorithm, and the third one uses the L* active learning algorithm. The proposed framework has been tested on various examples, among which a large case study of an autopilot coming from ADEPT, a toolset to support designers in the early design phases of automation interfaces. Experiences show that the method proposed in this thesis can identify existing automation surprises and also find new ones.
Pour plus d'informations à propos de la TVA et d'autres moyens de paiement, consultez la rubrique "Paiement & TVA".
Les commandes en ligne se font via notre partenaire i6doc.
Spécifications
- Éditeur
- Presses universitaires de Louvain
- Partie du titre
-
Numéro 459
- Auteur
- Sébastien Combéfis,
- Collection
- Thèses de l'École polytechnique de Louvain | n° 459
- Langue
- anglais
- Catégorie (éditeur)
- Sciences appliquées
- BISAC Subject Heading
- TEC000000 TECHNOLOGY & ENGINEERING
- Code publique Onix
- 06 Professionnel et académique
- CLIL (Version 2013-2019 )
- 3069 TECHNIQUES ET SCIENCES APPLIQUEES
- Description public visé
- Ingénieurs, ergonomistes
- Date de première publication du titre
- 16 janvier 2014
- Subject Scheme Identifier Code
- Classification thématique Thema: Technologie, ingénierie et agriculture, procédés industriels
- Type d'ouvrage
- Thèse
- Avec
- Index, Bibliographie
Livre broché
- Date de publication
- 12 octobre 2016
- ISBN-13
- 978-2-87558-494-6
- Ampleur
- Nombre de pages de contenu principal : 330
- Dépôt Légal
- D/2016/9964/27 Louvain-la-Neuve, Belgique
- Code interne
- 93852
- Format
- 17 x 24 cm
- Poids
- 559 grammes
- Type de packaging
- Aucun emballage extérieur
- Prix
- 25,10 €
- ONIX XML
- Version 2.1, Version 3
Google Livres Aperçu
Publier un commentaire sur cet ouvrage
Si vous avez une question, utilisez plutôt notre formulaire de contact
Sommaire
Résumé 11
Abstract 11
Remerciements 13
Introduction générale 17
Chapitre 1 : Le soucoupisme 25
1.1 La psychologie anomalistique 28
1.2 Le modèle sociopsychologique 32
1.3 Comment peut-on tester le modèle sociopsychologique ? 35
1.4 Les OVNI sont-ils un objet légitime de recherche ? 39
1.5 Nos hypothèses de travail 44
1.5.1 Il faut distinguer le débat exobiologique du débat ufologique 45
1.5.2 Le soucoupisme fait partie des nouvelles formes de religiosité 48
1.5.3 Les extraterrestres de l'ufologie sont une variation du schème culturel de la
présence d'une altérité parmi nous 49
1.6 Conclusion 53
Chapitre 2 : Les observations 57
2.1 Les origines du phénomène OVNI 60
2.1.1 Le folklore féérique 61
2.1.2 L'aviation 61
2.1.3 L’occulture 62
2.1.4 La science-fiction 62
2.2 Kenneth Arnold (1947) 63
2.3 Les méprises simples 64
2.4 Les méprises complexes 69
2.4.1 Les illusions 69
2.4.2 Les confabulations 70
2.4.3 La suggestibilité 73
2.5 Les hallucinations 75
2.6 Les faux souvenirs 80
2.7 Les mystifications 80
2.8 Conclusion 83
Chapitre 3 : Les enlèvements 89
3.1 L'affaire Betty et Barney Hill 92
3.2 Les thérapies Nouvel Âge 95
3.3 La personnalité encline à l'imagination 99
3.4 La paralysie du sommeil 101
3.5 Conclusion 102
Chapitre 4 : Les vagues 107
4.1 Premier contact 107
4.2 La Guerre des mondes 109
4.3 Panique ou illusion de masse ? 111
4.4 L’étude de Cantril et de ses assistants 116
4.5 L’inexplicabilité de la vague belge 121
4.6 L’explication sociopsychologique de la vague 128
4.7 Une illusion de masse engendrée par les médias ? 134
4.8 Conclusion 138
Chapitre 5 : Observation participante 143
5.1 L’enquête de terrain ufologique 146
5.2 Le chemin parcouru 149
5.3 La croyance au paranormal 156
5.4 Entretiens semi-structurés 159
5.4.1 Marc 163
5.4.2 Damien 165
5.4.3 André 171
5.4.4 Béatrice 176
5.5 Conclusion 178
Chapitre 6 : Science ou pseudo-science ? 183
6.1 Les critères externes 185
6.2 Les critères internes 189
6.3 Une science démocratique ? 193
6.4 Scepticisme ou pseudo-scepticisme ? 197
6.5 Conclusion 199
Conclusion 203
Références bibliographiques 215
Annexe 1 : Panel de discussions à Bruxelles Sceptiques au Pub 235
Annexe 2 : Entretien avec Patrick Maréchal 283
Annexe 3 : Entretien avec André 299
Annexe 4 : Entretien avec Damien 313