Redefining Question for Curve-Driving Practice Using Augmented Reality and Driving Models

Abstract

The purpose of this paper is to suggest a learning support system for curve driving that uses augmented reality (AR) and a model of curve driving. This learning environment focuses on the processes of recognition, judgment, and operation with an aim to enhance driving skills and tackle problems faced while learning how to drive. Driving entails an above cycle that is usually learnt through trial and error while it is practiced initially. Currently in Japan, the driving teacher provides instruction based on the accuracy of the learner’s operation, indicating that this learning method is focused on an ‘operation.’ When the teacher indicates an error in the learner’s operation, the learner has to consider the cause of the error by reviewing recognition, judgments, and operation from the teacher’s instructions. However, this method of consideration must be inferred and difficult for a novice driver. Moreover, the act of learning recognition, judgment, and operation separately is impossible because humans cannot divide them during driving. In this study, we developed a learning support system that can grasp recognition, judgment, and operation gradually and individually, based on driving models, AR for curve driving, scaffolding and fading, and reflective method. We designed a model of curve driving based on a situation awareness model. The learning environment implemented in this driving model can grasp recognition, judgment, and operation in fixed steps by information displayed in AR and can correct its own recognition, judgment, or operation by reviewing the result of each practice. This exercise will address the constraints and problems faced in tasks that involve learning curve driving; the question of the driving practice can be changed from “How do you drive a curve properly?” to “How do you perform recognition, judgment, and operation when you drive a curve properly?” We also conducted an experiment using the proposed learning support to test its effectiveness.