In general, most of the tasks performed by robots are related to picking up and placing objects, usually combined with trajectory planning and manipulation. For the objects to be picked up, they are given a generous collision detection volume until they finally approaches the target position. Occasionally, optimal trajectory may be too close to the obstacles to make the kinematics solver regard it safe and plausible. This situation gives us the intuition that the solver is much too conservative under certain collision avoidance constraints. Thus, our group worked with MoManTu and leveraged its state machine engine to test the successful rate for a simple picking and placement mission under in-door environment. During this mission, we made experiment about robotics navigation and arm manipulation based on inverse kinematics, and collect corresponding vedio results.