An Ambient Agent Model Exploiting Workflow-Based Reasoning to Recognize Task Progress

For an ambient intelligent agent to support a human in demanding tasks it is important to be aware of the progress made in a given workflow. It would be possible to interact with the human about this, but this may disturb task performance. In this paper an ambient agent model is presented that is able to obtain such an awareness of the human's progress in a workflow by performing model-based analysis using available workflow models and available observation information. Simulation experiments for the presented the ambient agent model for a case study are discussed, and evaluated by automated formal verification. Keywords. Ambient agent, workflow model, task progress, awareness, modelbased reasoning.

• Type: Conference
• Year: 2008
• Where: AMI
• Authors: Fiemke Both, Mark Hoogendoorn, Jan Treur
• Keywords: Ambient Agent, Ambient Agent Model, Ambient Intelligent Agent, AMI 2008, Artificial Intelligence