This paper addresses the challenging task of predicting lane changes for remote vehicles approaching the ego vehicle from two lanes away in a highway driving scenario. Existing systems, such as the Blind Spot Indicator (BSI), detect and notify the ego vehicle's driver when a vehicle is merging adjacent to it, but these notifications are triggered only when the merging vehicle is within a pre-defined distance. However, this approach falls short when both the ego and remote vehicles are simultaneously merging into the middle lane from opposite sides, as the notification is given too late for the ego vehicle's driver to respond appropriately. The primary focus of this study is to develop a predictive model capable of discerning lane change behavior in real-time for remote vehicles approaching two lanes away. Specifically, the research concentrates on a scenario involving three vehicles on a 3-lane road. The ego vehicle (EV), situated on the left lane, aims to merge into the empty middle lane, while the lead vehicle (LV) and the target vehicle (TV) are positioned on the right lane. The TV is passing the LV and moving towards the middle lane. The authors developed an SVM-based model to predict, with 82% accuracy, in real-time, a lane change behavior. This allows timely notifications to the EV driver, preventing simultaneous lane changes and reducing the risk of side collisions. To determine the appropriate timing for notifications, a preliminary user study using vehicle simulation was conducted.