Fuzzy-based Approach for Road Accident Risk Estimation on GPU

Understanding road accidents is very important nowadays, especially if self-driving is the long-term technological goal. However, uncovering the causes of a particular accident is a very complex issue, so applying fuzzy methods can help solve the problem. In most cases, the accident is caused by human mistakes, such as speeding, but the environmental conditions must be examined regardless. The curve and the slope of the road, as well as the weather conditions, can cause the accident, so in many cases, it can be more useful if the route is planned based on the characteristics of the road in certain cases. This article presents the fuzzy-based risk estimation of certain properties of road features and possible parallelism possibilities. In addition, statistics on the environmental conditions of individual accidents in Hungary and their conclusions will be presented.