Project

Back to overview

Designing automation for visual inspection tasks: influence of support system properties and work design factors

English title Designing automation for visual inspection tasks: influence of support system properties and work design factors
Applicant Sauer Jürgen
Number 149184
Funding scheme Project funding (Div. I-III)
Research institution Departement für Psychologie Universität Freiburg
Institution of higher education University of Fribourg - FR
Main discipline Psychology
Start/End 01.03.2014 - 31.08.2018
Approved amount 293'703.00
Show all

Keywords (3)

work design; adaptable automation; visual inspection

Lay Summary (French)

Lead
L’inspection visuelle, consistant à identifier des irrégularités dans des objets, est au cœur de nombreuses activités professionnelles comme le diagnostic oncologique, le contrôle de qualité des produits ou la sécurité. Ne pas détecter l’une d’elles peut avoir des conséquences négatives importantes, comme une tumeur maligne non détectée, un circuit imprimé défectueux utilisé dans l’assemblage du système de bord d’une navette spatiale, ou un engin explosif embarqué à bord d’un avion.
Lay summary

L’un des moyens pour soutenir les professionnels dans cette activité complexe, parfois fastidieuse et exigeant une attention constante sur de longues périodes pourrait se trouver dans l’introduction d’un système de détection automatique d’objets suspicieux. Ce projet vise à observer dans le contexte de la détection d’objets prohibés dans les bagages à main non seulement la pertinence d’un tel système mais aussi l’influence de son design, ainsi que d’autres facteurs liés aux conditions de travail, sur la performance de l’utilisateur.

Dans un premier temps, cette recherche se concentrera sur l’efficacité de l’automatisation adaptable, un système où l’utilisateur peut lui-même décider à chaque instant du niveau de support, par rapport à une forme plus conventionnelle d’automatisation statique où le niveau d’aide est prédéfini et ne peut être changé. Dans un deuxième temps, le projet se penchera sur l’influence des caractéristiques spécifiques de l’algorithme de détection (par exemple sa fiabilité) sur la façon dont les opérateurs utilisent le système ainsi que la pertinence de familiariser les utilisateurs à ces caractéristiques durant la phase d’entrainement. Finalement, la pertinence du système de détection automatique sera confrontée à deux aspects importants du travail lié à la sécurité dans un aéroport, à savoir la segmentation du temps de travail et le niveau d’expertise.

Cette étude intègre de nombreux aspects psychologiques rencontrés sur le lieu de travail, tels que l’interaction entre l’humain et la machine (ergonomie cognitive), l’expertise et la segmentation du temps de travail (psychologie du travail) mais aussi l’attention visuelle (psychologie cognitive). Elle contribue à évaluer le bénéfice de l’automatisation dans le cadre de l’inspection visuelle et à en adapter le design en fonction des caractéristiques de l’utilisateur.
Direct link to Lay Summary Last update: 08.10.2013

Responsible applicant and co-applicants

Employees

Publications

Publication
Automation in visual inspection tasks: X-ray luggage screening supported by a system of direct, indirect or adaptable cueing with low and high system reliability
Chavaillaz Alain, Schwaninger Adrian, Michel Stefan, Sauer Juergen (2018), Automation in visual inspection tasks: X-ray luggage screening supported by a system of direct, indirect or adaptable cueing with low and high system reliability, in Ergonomics, 1-35.

Collaboration

Group / person Country
Types of collaboration
Prof. Adrian Schaninger (Hochschule für Angewandte Psychology, Fachhochschule Nordwestschweiz) Switzerland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Prof. Jason McCarley (Flinders University, Adelaide) Australia (Oceania)
- Research Infrastructure

Scientific events

Active participation

Title Type of contribution Title of article or contribution Date Place Persons involved
Annual Meeting of the Human Factors and Ergonomics Society Europe Chapter Poster Visual inspection of baggage X-ray images: How do expertise and automation influence detection performance? 08.10.2018 Berlin, Germany Chavaillaz Alain;
Annual Meeting of the Human Factors and Ergonomics Society Europe Chapter Talk given at a conference Increasing operator flexibility in luggage screening: The impact of spontaneous breaks and adaptable automation on performance 28.09.2017 Rom, Italy Chavaillaz Alain;
10. Tagung der Fachgruppe Arbeits-, Organisations- und Wirtschaftspsychologie der Deutschen Gesellschaft für Psychologie Talk given at a conference The impact of work-rest schedules and adaptable automation on airport security screener’s performance 13.09.2017 Dresden, Germany Chavaillaz Alain;
15th Congress of Swiss Society of Psychology Talk given at a conference Operator autonomy in luggage inspection: Effects on performance 04.09.2017 Lausanne, Switzerland Chavaillaz Alain;
14th Congress of Swiss Society of Psychology Talk given at a conference Automatic support in luggage screening: effects of system reliability and degree of automation on performance. 08.09.2015 Geneva, Switzerland Chavaillaz Alain;


Associated projects

Number Title Start Funding scheme
134566 Effectiveness of explicit and implicit adaptive automation under different work stressors 01.09.2011 Project funding (Div. I-III)
188808 Improving work design for airport security officers 01.11.2019 Project funding (Div. I-III)

Abstract

The present project is concerned with the design of automation to support operators in visual inspection tasks. While these issues have attracted increasing interest in psychological research (notably in quality control in manufacturing, medical diagnoses, and airport security), the potential benefits of automation in visual inspection may not have been fully reaped. The proposed project relates to two preceding SNSF-projects which examined the effectiveness of different forms of automation in the application of process control. It would allow us to determine to what extent the findings gained in highly complex multiple-task environment are transferable to the domain of visual inspection, which is characterised by a single task with considerable mental demands.The present project pursues two main goals. First, it aims to examine whether modern concepts of automation design (e.g., adaptable automation) can also be introduced to a domain requiring high levels of vigilance, like luggage screening, and to determine whether they are of similar benefit in that domain as they have been in others. Second, it aims to examine the implications of introducing higher levels of automation for other work design factors in visual inspection, such as training and rest breaks. These factors are an integral part of the overall socio-technical system and their impact need to be tested in multi-factorial experimental designs.Prior to the completion of the empirical work, a simulation will be developed to model the work environment of luggage screening officers. This simulation environment will be able to model different automation modes (e.g. no automatic support, system informs about presence of target, system indicates location of target) and will be suitable for use in lab-based work as well as in field research. Five experiments are envisaged to examine the effects of providing different forms of automatic support to the operator, in combination with pertinent work design issues surrounding visual inspection tasks (e.g. rest break, training). Experiment 1 is concerned with the effectiveness of adaptable automation (i.e. operators are free to select their preferred level of automation and change it at any time) compared to the more conventional form of static automation (i.e. level of automation is predetermined by the system). Experiments 2 and 3 examine the effects of changing target detection algorithms (e.g., increasing its sensitivity) under different levels of system reliability but also examining the potential of training the operator on automation characteristics. Experiments 4 and 5 adopt a broader perspective by including other important aspects of work design for airport security, such as work/rest schedules and expertise. While a central outcome variable of the research refers to operator performance (notably accuracy and speed in target detection), the measurement of eye movements will provide data about the operators’ visual search behaviour as an important complementary measure. Further measures include automation reliance, trust in automation, and subjective workload.The planned work is expected to provide a number of benefits. First, a task environment will be developed which can be used for lab-based work to model the technical work environment of luggage screening officers. The task environment will be made available to the research community. Second, the findings are expected to contribute to the design of support systems in the context of airport security and other visual inspection tasks.
-