Digital Twin for Simulation and Evaluation of Assistive Navigation Systems

Abstract

The assistive navigation of visually impaired individuals requires the development of different algorithms for obstacle detection, recognition, avoidance, and path planning. The assessment and optimization of such algorithms in the real world is a painstaking process that requires repetitive measurements under stable conditions, which is usually difficult to achieve and costly. To this end, digital twin environments can be used to replicate relevant real-life situations, enabling the evaluation and optimization of algorithms through adjustable and cost-effective simulations. This chapter presents a digital twin framework for the simulation and evaluation of assistive navigation systems, and its application in the context of a camera-based wearable system for visually impaired individuals in an outdoor cultural space. The system incorporates an obstacle avoidance algorithm based on fuzzy logic. The utility and the effectiveness of this framework are demonstrated with an indicative simulation study. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.