Real-Time Traffic Risk Detection Model Using Smart Mobile Device

Abstract

Automatically recognizing dangerous situations for a vehicle and quickly sharing this information with nearby vehicles is the most essential technology for road safety. In this paper, we propose a real-time deceleration pattern-based traffic risk detection system using smart mobile devices. Our system detects a dangerous situation through machine learning on the deceleration patterns of a driver by considering the vehicle’s headway distance. In order to estimate the vehicle’s headway distance, we introduce a practical vehicle detection method that exploits the shadows on the road and the taillights of the vehicle. For deceleration pattern analysis, the proposed system leverages three machine learning models: neural network, random forest, and clustering. Based on these learning models, we propose two types of decision models to make the final decisions on dangerous situations, and suggest three types of improvements to continuously enhance the traffic risk detection model. Finally, we analyze the accuracy of the proposed model based on actual driving data collected by driving on Seoul city roadways and the Gyeongbu expressway. We also propose an optimal solution for traffic risk detection by analyzing the performance between the proposed decision models and the improvement techniques.

Exploiting Multi-modal Contextual Sensing for City-bus's Stay Location Characterization: Towards Sub-60 Seconds Accurate Arrival Time Prediction

“…Hoang [16] utilizes GSM radio signal, road network data, and weather data in addition to GPS logs to capture seasonal changes and instantaneous changes. Park et al [26] have used the GPS data in combination with camera images to propose a traffic risk detection model which automatically detects dangerous driving situations by monitoring the driving behavior. Vij et al [43] have effectively used microphone data of smartphones to identify different traffic states.…”

Analysis of Machine Learning Techniques Applied to Sensory Detection of Vehicles in Intelligent Crosswalks

“…AI techniques—and ML in particular—also enable the development of smart road safety (SRS) solutions for smart cities. In [ 25 ], for instance, a rear collision detection system for drivers was modelled using the vehicle acceleration and distance from the preceding car, and both random forest and neural networks were used. For the detection of vehicles moving in the wrong direction along highways, a camera-based system classifying the direction of the vehicle was proposed in [ 26 ], allowing traffic authority to be notified and act against such dangerous situations.…”

RoadSpeedSense: Context-Aware Speed Profiling from Smart-phone Sensors

“…Hoang [6] et al utilizes GSM radio signal, road network data, and weather data in addition to GPS logs to capture seasonal changes and instantaneous changes. Park et al [13] have used the GPS data in combination with camera images to propose a traffic risk detection model that automatically detects dangerous situations by monitoring driving behavior. Vij et al [21] have used microphone data of smartphones in an effective way of identifying different traffic states.…”

“…Existing Works: Prior research has been carried out on intelligent road traffic and transport management systems [3,23] road traffic congestion measurement [12], travel time estimation [20], Modelling urban mobility pattern, Characterization of city traffic and road, etc. Researchers have worked on off-line or online GPS trace of vehicles, online social media data [22], sensor data from accelerometer and microphone of highend mobile phone or image or video captured by roadside cameras [13]. GPS trace is analyzed to develop transport and traffic monitoring systems, traffic congestion, or any special event is extracted from online social media.…”

“…11 https://www.ncbi.nlm.nih.gov/pmc/articles/ 12 https://developers.google.com/maps/documentati on/distance-matrix/start 13 https://stackoverflow.com/questions/432773 42/can-we-get-speed-limit-information-fromwaze-apis 14 http://opentraffic.io/#partes 15 Space-mean speed is the harmonic mean of the spot speed over a specified road segment of unit length. Space mean speed is always lower than the time mean speed as it weights slower vehicles more heavily as they occupy the road stretch for longer duration of time Contribution: Firstly, through an extensive study, we identified features that significantly influence the aggregate speed profile of a road segment and also can be easily derived.…”