With rapid advancements in technology, new devices such as interactive voice systems, navigation systems, hands-free mobile communications, and entertainment systems have introduced a significant range of user controls/demands within vehicles. Even though these applications assist the multi-tasking ability for the driver, they introduce a variety of distractions that divert the driver's attention from the primary driving task. Although there has been significant advancements in active in-vehicular safety systems, the number of accidents, injury severity levels and fatalities has not reduced. In fact, human error, low performance, drowsiness and distraction may account for a majority of accidents. Current active safety systems utilize the vehicle dynamics and are unaware of context and driver status, so they do not adapt to changing mental and physical conditions of the driver. This study proposes an active safety system structure as a first step in realizing robust, human-centric and intelligent active safety systems with the ability to adapt/reconfigure themselves to the context and driver status. This work develops, evaluates and combines three sub-modules: biometric driver identification, maneuver recognition, and distraction detection systems. It also attempts to classify distraction into levels (i.e., no, low, medium, and high). The proposed system contributes in four areas: (1) Achieve Robust Identification, (2) Prune the available information space by personalizing maneuver recognition and distraction detection, (3) Improve response time and performance by predicting the driver's distracted behavior for possible use in accident prevention, (4) Provide recommendations for evaluating risk factors by classifying distraction. Overall system performance is evaluated on the UTDrive Corpus, confirming the suitability for real-world applications.