Abstract:
Most organizations are realizing the value of data; and Business Intelligence (BI) is the key to analyze data in today’s information world. BI is used to investigate real world scenarios to identify potential problems. One such scenario is the ever increasing rate of auto-mobile accidents in Pakistan. Most of the countries are gaining benefits by analyzing the past road accidents data and forecasting a lot of information about possible future accidents. But unfortunately in Pakistan, data is collected to a limited extent, and the standard format of collecting data is not truly followed. There is no central department for accident prevention neither is there any data repository. Hence the number of accidents is steadily on the rise.
The aim of this project is to improve road safety and reduce accidents by identifying the root causes of these accidents. In order to achieve this goal, we have analyzed the traffic accidents data of Pakistan’s Motor-ways and National highways. The data was collected from National Highways and Motorway Police, cleaned and modeled on different dimensions for analysis and prediction. The major recommendations concluded from the results are as follows.
1. Sunder (near Lahore) had the most number of accidents due to non-availability of pedestrian overhead bridge; and over speeding of drivers. A solution is to install overhead bridge and to have police patrol the area round the clock to keep drivers in check.
2. Trucks and trollies are involved in the most dozing at the wheel accidents; hence we need to plant road bumps on locations such as Sahiwal, where these accidents are more prominent.
3. Motorcycle riders have the most frequent fatal accidents among all vehicle types; the reason being disregard of safety by not wearing the helmet. A solution is to impose heavy fines on driving without helmet.
