The Virtual Driving Instructor: A Multi-Agent Based System for Driving Instruction

Abstract

Driving simulators need an Intelligent Tutoring System (ITS). Simulators provide ways to conduct objective measurements on students' driving behavior and opportunities for creating the best possible learning environment. The generated traffic situations can be influenced directly according to the needs of the student. We created an ITS -the Virtual Driving Instructor (VDI) -for guiding the learning process of driving. The VDI is a multi-agent system that provides low cost and integrated controlling functionality to tutor students and create the best training situations. Its architecture offers flexibility for adding and extending several awareness types that are required for a good driving instructor. In our design, we added situational, adviser, presentation and curriculum awareness. The design provides a simple interface for future extensions, such as the addition of student, environmental and scenario awareness. AbstractDriving simulators need an Intelligent Tutoring System (ITS). Simulators provide ways to conduct objective measurements on students' driving behavior and opportunities for creating the best possible learning environment. The generated traffic situations can be influenced directly according to the needs of the student. We created an ITS -the Virtual Driving Instructor (VDI) -for guiding the learning process of driving. The VDI is a multi-agent system that provides low cost and integrated controlling functionality to tutor students and create the best training situations. Its architecture offers flexibility for adding and extending several awareness types that are required for a good driving instructor. In our design, we added situational, adviser, presentation and curriculum awareness. The design provides a simple interface for future extensions, such as the addition of student, environmental and scenario awareness.