Sampling-based collision warning system with smartphone in cloud computing environment

Abstract

For improvement of road safety, many collision-warning systems are developed. In this study, we propose Sampling-based Collision Warning System (SCWS) that overcomes the limitations of existing collision warning systems such as high installation cost, requirement of high market penetration rate, and the lack of consideration of traffic dynamics. SCWS gathers vehicle operation data though smartphones of drivers on the road and shares the information of surrounding vehicles' movement through a cloud server. From the pool of information on the cloud, SCWS uses sampled data, which indirectly represents the traffic state and traffic changes in the perspective of the leader vehicle. Therefore, SCWS can effectively replace the leader vehicle's information with the average behavior of sampled surrounding vehicles. The performance of SCWS is evaluated with comparison to Vehicle-to-Vehicle communication based Collision Warning System (VCWS) and Infrastructure based Collision Warning System (ICWS), where VCWS is considered the most similar measure to the actual collision risk in theory, but in practice very difficult to achieve due many limitations, such as high installation cost and market penetration. The result shows that in both aggregation and disaggregation level analysis the proposed SCWS exhibits a similar collision risk trend to the VCWS. Furthermore, the SCWS shows a high potential for practical application because it has the acceptable performance even with a low sampling ratio (40%), requiring a low market penetration rate and low installation cost by using the wide spread smartphone.