Most conventional path generation algorithms search for an optimal path that avoids collisions with obstacles under the constraint of platforms' kino-dynamics. These conventional algorithms usually assume that a user position and obstacle locations are accurately known at any point in navigation environments. However, in a positioning network, the accuracy of a position estimate varies depending on, e.g., the ranging accuracy, network geometry, multipath error, and signal blockages which may lead to unexpected situations, including collision and low efficiency path planning. Therefore, positioning accuracy must be considered in path generation to ensure a reliable navigation capability and collision avoidance. To consider positioning accuracy in path planning, the proposed method in this paper uses a mixture of potential and positioning risk fields that generates a hybrid directional flow to guide an unmanned vehicle (UV) in a safe and efficient path. The results of simulations showed that the proposed method generated successful paths for around 90% percent of the tested routes, while using only the potential field method failed for around 50%. To demonstrate the effectiveness of the proposed local hybrid path planning method, we perform an experiment using a small-size quadcopter, and the results are analyzed and discussed.