According to World Health Organization 2015 report, India has one of the highest number (137, 572) of traffic accident fatalities when compared to other countries. There is a need to identify the accident characteristics and implement cost effective safety measures to reduce the fatalities. The objective of this study is to prioritize the issues related to Indian road accidents using MORTH and ADAC, NATRIP databases based on a triple-layer approach (i) “Society” related to infra-structure (ii) “Individual” related to drivers and (iii) “Vehicle” related to vehicle safety regulations. At first, macro analysis was performed on MORTH 2014 data and then followed by micro analysis was carried out with ADAC, NATRIP data (National Highway-08). In order to identify the characteristics for those NH-08 accidents, a data mining technique, Self-Organizing Maps (SOM) was applied. Macro-analysis results show that 45% of fatalities are related to society (infrastructure), 29% are related to individuals (driver) and 26% are related to vehicles (rules and regulations). Micro-accident analysis results indicate that more signal availability (30%) in urban areas compared to that in rural areas leads to less intersection accidents. Improper driving-manner (30%), improper lane-change (16%), failing to use restraint system (29%) are the important elements related to individual. Rear crashes (54%), angled/side crashes at junctions (14%), accidents caused by trucks (48%), and hitting stranded parked-vehicles (24%) are the influential factors related to vehicles. Five basic traffic injury patterns are revealed by SOM analysis within ADAC, NATRIP data and a few countermeasures were proposed for each group separately.
Chinmoy Pal, Takahashi Nobuhiko, Sangolla Narahari, Manoharan Jeyabharath and Tajima Azumi