Intelligent Transportation Systems (ITS) demand driving safety as an eminent design requirement for future generation vehicles. Collision evasion as well as consequent casualties minimization command timely delivery of significant precautionary information to the drivers. Consequently, the driver may get a clear view about the present driving situation and be able to adopt timely decision to circumvent the forthcoming dangers. This research work proposes an Adaptive Neuro-Fuzzy Inference System (ANFIS) based situation assessment method that supports the drivers to take up suitable decisions by analyzing the driver behavior of preceding/succeeding cars. The proposed approach models the stability of drivers in the perspective of connected cars and deduce the current stability situations from the sensors which are implanted in the cars. This connected cars scenario for Collision Warning System (CWS) is simulated using three Raspberry Pi boards along with ultrasonic sensor, gas sensor and accelerometer sensor. These sensor data are transmitted to other preceding or succeeding cars using visible light communication. Subsequently, these data are processed using both Mamdani and ANFIS model for situation assessment which provides the stability level of drivers. The result concludes though the Mamdani model quickly computes the stability of driver by analyzing the sensor data, it suffers from low sensitivity and precision when compared to ANFIS which showcases higher sensitivity and precision. © Springer International Publishing AG 2018.