Driving Analytics for Improved Road Safety
MetadataShow full item record
Road safety has always been a globally growing concern and speeding is one of the major factors that causes road crashes. Road geometry is an important factor that influence speeding and thus the road safety. The purpose of this research is to access the sensitivity of horizontal and vertical road geometries on driving safety. A simulation framework was developed to imitate a safe human way of driving along 2-lane rural roads during free flow or no traffic conditions. This framework is in a computer simulation environment to generate a safe driving speed profile by using road geometry information. The framework also includes the real-time importing of the required road geometry data from the online map databases like Open Street Maps and Google Maps. Basically, if the model is inputted with latitude and longitude coordinates of starting and end point for a route, the model will output for every 1 meter along the route the simulated driving speed under a stipulated safe driving condition. Based on the starting and the end coordinates of a driving route, the model queries the coordinates and the elevation of the equidistant waypoints along the route from online map databases. This queried information about the road geometry is used to evaluate the safe driving speed conditions along the route. The model tries to imitate the human way of driving by predicting at each point along the route if a driver will accelerate, maintain a constant speed, decelerate without braking or apply brake to stay within the limiting speeds due to road geometry. The simulation framework was validated against the real driving speed profile recorded on four routes. As an improvement to road safety, this framework could be deployed to warn drivers when they are having unsafe driving speeds.
Ajith, Adithya (2018). Driving Analytics for Improved Road Safety. Master's thesis, Texas A & M University. Available electronically from