Algorithm Development for AI Driven Path Planning Using Augmented Reality for Navigating in Indian Complex Road Scenarios

The road infrastructure in India has complex navigational challenges with most of the road unstructured especially in rural areas. Decision-making becomes a challenge for drivers in unpredictable environments such as narrow roads, flooded roads and heavy traffic. In this paper, an Augmented Reality based ML-Algorithm for Driver Assistance (ARMADA) has been proposed that improves awareness to safely maneuver in these conditions. The methodology for development and validation of this Augmented Reality (AR) based algorithm contains multiple steps. Firstly, extensive data collection is conducted using real time recording and benchmark datasets like Berkeley Deep Drive (BDD) and Indian Driving Dataset (IDD). Secondly, collected data are annotated and trained using an optimal machine learning (ML) model to accurately identify the complex scenario. In third step, an ARMADA algorithm is developed, integrating these models to estimate road widths, detect floods and provide seamless driver assistance in a Human Machine Interface (HMI). Finally, proposed algorithm undergoes validation to ensure its effectiveness and accuracy in real-world practical scenarios. The result of this study concludes significant improvements in driver decision making and safety of the driver in complex maneuvering.