Route planning, which is used to calculate feasible routes in a given environment, is one of the key issues in navigation systems. According to different constraints in different given space, various route planning strategies have been developed in recent years. Current route planning models for indoor space focus on providing routes for pedestrians or fix-sized users, like robots and persons in wheelchairs. None of the existing model can provide feasible routes for arbitrary-shape users, which appears to be more and more useful in many situations, like users driving small indoor autos or moving carts with products. This paper proposes a two-phase route planning model which can support route planning for users with arbitrary shapes. In the first phase, the LEGO model represents the entire space by using different types of cubes. These cubes are further merged in the second phase to form the maximum accessible blocks. By computing the maximum accessible widths and lengths between b...