Alerting for vehicles demonstrating hazardous driving behavior

Abstract

Cooperative Collision Warning Systems (CCWSs) have become a major vehicle safety application in intelligent transportation systems. Vehicles organized in a vehicular ad-hoc network use a CCWS communication protocol to propagate emergency messages about hazardous events. Police cars, ambulances responding to incidents and speeding cars or motorcycles that constantly vary their speed, change lanes or commit other apparent traffic violations are examples of vehicles that demonstrate hazardous traffic patterns. Using their GPS and motion sensors, vehicles can detect those traveling in nearby avenue sections who constitute a threat.In this paper, we propose a broadcasting protocol that alerts drivers about the presence of moving vehicles demonstrating hazardous driving behavior. In order to limit the volume of redundant transmissions, our approach selects the vehicles to be responsible for transmitting the emergency information for a hazardous vehicle. In this context, we provide mechanisms to create and maintain a chain of transmitters. This chain "covers" the road sections on which a hazardous vehicle is moving.Our protocol attempts to increase the probability that an endangered vehicle does obtain timely information about a hazardous vehicle and reduce the total communication traffic imposed in urban environments where the vehicles' density is often high. We experimentally evaluate our suggested protocol by comparing it with two alternative CCWS broadcasting approaches and we ascertain the extent in which the above objectives are met.

Acceptability Evaluation of Inter-driver Interaction via a Vehicle Agent Using Vehicle-to-Vehicle Communication on a Driving Simulator

“…A study by the National Highway Traffic Safety Administration (NHTSA) showed that V2V communication can reduce up to 80% of crashes [2]. Over the last decade, many researchers have proposed protocols and concepts to alert drivers of upcoming potential hazards by exchanging information with other vehicles using V2V communication [3], [4], [5], [6]. Based on information on the motion of neighboring vehicles, Sengupta et al presented the concept of a cooperative collision warning system that provides warnings and situation awareness displays to drivers [3].…”

“…Makanae et al proposed a prototype that utilizes a wireless local area network and a GPS to send messages to vehicles around a vehicle [17]. Verroios et al proposed a method for limiting the range of communication by utilizing a motion sensor and a GPS [4]. Moreover, techniques have been proposed to identify which vehicle communicates with one's vehicle by combining the vehicle position from V2V with the relative position and features from sensors such as an image sensor and a radar [18].…”

“…Currently, most existing and potential V2V communication services involve notifying drivers of an increased risk of collision with another vehicle when changing position or speed (e.g., intersection movement assist [3], [4], [5], [6]). However, in actual traffic environments, drivers often use multiple modalities (e.g., gestures, headlight flashing, and horns) to warn other drivers while negotiating a potentially risky maneuver [7], [8].…”

Highly Robust Clustering of GPS Driver Data for Energy Efficient Driving Style Modelling

“…While the authors of [4] use statistical measures, the authors of [14] combine a clustering algorithm with complex neuroscale and Bayesian factor analyses. A similar research is also performed in [16], where cars are analysed for potentially dangerous behaviour. However, the focus of the latter work lies more on the communication protocols between the traffic participants and not on their classification.…”