A review on multiple facets of Indian Road safety-classification, challenges and future scope

Road safety as a domain is not limited to crash prediction but has multiple facets like- (1)Intelligent Driving, (2)Safety systems for vehicles, (3)Advanced Crash prediction with warning system, (4)Driver behaviour monitoring & mapping, (5)Traffic analysis, (6)Traffic sign identification, (7)Traffic monitoring & mapping, (8)Work zone safety, (9)Time dependent traffic analysis, (10)Vehicle orientation detection, (11)Pedestrian and zonal safety and (12)Driver assistance among others. This study primarily introduces the Road safety domain with appropriate data backed background elucidation. Then it classifies various area of the domain namely- Crash Prediction, Crack, pothole, pavement hole, quality of road detection, Detection of objects (station & moving), Special Use cases (Helmet Detection, Driver behavior-environment mapping ,Orientation detection model and Road marking classification models). Finally it presents comprehensive review along with some future trends in the research domain via tabular method for optimized comparative analysis along with a brief description of Indian Road safety scenario. This paper aims to provide more attention of society and researchers to Road safety and various facets that directly influence human life, Road safety being prominent reason for road kill.

I. Introduction

Road Safety around the globe has become an issue of significant importance ever the since the wave of industrialization hit the world economy, leading to massive industrialization, development of new roads, a huge surge of vehicles, exponential development of transport & tourism industry. This growth is even more significant in developing countries like India, leading to a surge of road rage, road accidents and development of a road safety architecture. It results in millions of death each year as accumulated by Ministry of road safety & highways [1] [2], even after implementation of strict human surveillance to ensure road safety measures are followed. Multiple measures are taken by the Ministry of road safety & highways namely [3], (1) Establish a Road Safety Information Database, (2) Ensure safer Road infrastructure, (3) safer drivers, safer vehicles (4) Safety and marking of accident-prone area, (5) Road safety education and training, (6) Enforcement of safety laws, (7) Emergency Medical Services for Road Accidents, (8) HRD & Research for Road Safety, (9) Strengthening Enabling Legal, Institutional and Financial Environment for Road Safety and (10) Implementation Strategy. But deliberating the data from previous decade to latest available data as facilitated by Ministry of road safety & highways of India [4][5], despite above strategies, repetitive iterations of measures, intense manual surveillance and extensive pecuniary advances in the same, the number of road accidents has been increasing with a steady pace. This indicates the need for technical intervention, implementing concepts of predictive modelling, image segmentation, deep learning & machine learning frameworks. They can deliberate multiple attributes like crash data, the geometric and operational characteristics of the road, and the environmental conditions, incorporating the most important factors [6]. Variegated modelling techniques have been developed over the period that improves representation, analysis & prediction on grounds of road safety.