A Computationally-Efficient Collision Early Warning System for Vehicles, Pedestrians, and Bicyclists

This paper describes a computational architecture of a collision early warning system used for vehicles and other principals such as pedestrians and bicyclists. Early warnings allow drivers to make good judgments, and to drive proactively and defensively, in order to avoid emergency stopping or dangerous maneuvering. The authors describe how the presence of many principals in a dense intersection, it is difficult to predict even a few seconds in advance, since there are an enormous number of possible scenarios. They propose a two-stage collision risk assessment process that is designed to track only the more plausible collisions and, of those, to alert drivers only to the most critical situations. The collision risk assessment process applies a library of scenarios, first using highly efficient geometric comparisons, and then selectively applying more elaborate statistical analysis to compute an assessment of potential accidents that contains enough information for the user interface to effectively decide if, when, and how to warn the driver. The authors conclude that this two-stage assessment is computationally efficient and can work in real-time for complex intersections.