While many efforts are currently devoted by research bodies to investigate in robot harvesting, challenges related to picking fruits from clusters are still considered an open issue which can limit the operation success. On the other hand, existing planning frameworks for robotic manipulation in cluttered and uncertain environment are getting more and more attention for their ability to deal with physics-based strategies to free the robot path to a goal object. However, those approaches are either computationally expensive and/or designed for 2-D occlusion scenes. Consequently, they are not readily applicable to the complex 3-D geometry of fruits in clusters. In this work, we present a path planning algorithm for pushing occluding fruits to reach-and-pick a ripe one. Hence, we propose an Interactive Probabilistic Movement Primitives (I-ProMP) which is computationally efficient and is readily used for 3-D problems. We demonstrate the efficiency of our approach with pushing unripe strawberries in a simulated polytunnel. Our experimental results confirm I- ProMP successfully pushes table top grown strawberries and reaches a ripe one.