A computationally-efficient collision early warning system for vehicles, pedestrians and bicyclists
This paper describes the computational architecture of a collision early warning system for vehicles and pedestrians. Early warnings allow drivers to make good judgments and to avoid emergency stopping or dangerous maneuvering but poses a difficult computational problem. With many principals (vehicles, pedestrians, bicyclists, etc) coexisting in a dense intersection, it is difficult to predict even a few seconds in advance, since there are an enormous number of possible scenarios and potential collisions. It is a major challenge to manage computational resources and human resources so that only the more plausible collisions are tracked and of those, only the most critical collisions prompt warnings to drivers. In this paper, we propose a two-stage collision risk assessment process, consisting of (1) a preliminary assessor which throughly considers all surrounding principals and identifies likely potential accidents via simple efficient geometric computations, and (2) a specialized assessor which focus on the more plausible accidents and computes more accurate collision probabilities via sophisticated statistical inference. The whole process delivers an expected utility assessment to available user-interfaces, allowing the user interfaces make discriminating choices of when to warn drivers or other principals.
Greene, D. H. ; Liu, J. J. ; Reich, J. E. ; Hirokawa, Y.; Mikami, T.; Ito, H.; Shinagawa, A. A computationally-efficient collision early warning system for vehicles, pedestrian and bicyclists . Fifteenth World Congress on ITS; 2008 November 16-20; New York, NY.