A Sampling Model to Ascertain Automation-Induced Complacency in Multi-Task Environments

: This article discusses the development of a model that defines the optimal sampling behaviour of operators in a multi-task flight simulation, where one of the tasks is automated. The goal of this model is to assign a cost function to the attention allocation strategy of participants, allowing us to assess the efficiency of their overall strategy. The model revealed that the optimal sampling strategy should be the same regardless of the automation reliability. When applied to previously reported empirical data, the model revealed that participants using constant, highly reliable automation demonstrated more `expensive' monitoring behaviour. However, their monitoring behaviour became more efficient over time, which is inconsistent with the conclusion that the poor overall monitoring performance was due to complacency. This model allowed us to define an optimal monitoring performance, which is an important step in being able to accurately assess "complacency". Key words: ...

• Type: Conference
• Year: 2004
• Where: IFIP13
• Authors: Nasrine Bagheri, Greg A. Jamieson
• Keywords: Artificial Intelligence, IFIP13 2004, Model, Optimal Sampling, Optimal Sampling Behaviour