Indoor fire environments with smoke, restricted spaces, and obstacles bring challenges to the path planning of firefighting robots. This paper proposes a two-layer path planning method. First, a map segmentation algorithm is designed to segment indoor environments into an undirected graph with multiple rooms connected by doors and performs the shortest path search. Then, an ellipsoid constraint-based bat algorithmic planner (EBAT) for indoor environments is proposed. Simulation results verify the effectiveness of the proposed method for firefighting robot path planning.