Abstract: Soft Computing can be used for the analysis of driverís performance which may help the licensing authorities to make a better judgment about the skills of a driver. We have used Fuzzy Logic in the proposed system. The readings (engine rpm, engine load, vehicle speed and acceleration) are obtained from the car using sensors and are used as input for the fuzzy system. Fuzzy C-Means Clustering (FCM) with unsupervised learning is used for obtaining input membership functions. The fuzzified parameters are fed into the Fuzzy Inference System (FIS) along with the rule base to obtain a crisp output as a final result. The crisp output gives degree of correctness along with a report on driving skills of a driver.

Keywords: Fuzzy C-Means Clustering, Membership function, Fuzzy Inference System, fuzzification, unsupervised learning.