Energy-Efficient Shortest Path Planning on Uneven Terrains: A Composite Routing Metric Approach

This work investigates the problem of energy-efficient shortest path planning for autonomous unmanned vehicles (AUVs) in outdoor natural terrain. We use an established energy-cost model for mobile robots, which integrates the terrain's angle of declination with the gravity and wheel-terrain friction forces, to calculate the AUV's energy consumption. We propose a novel composite routing metric that combines the energy consumption along the path with the traveled distance, to create a variant of the famous Dijkstra algorithm for path planning. Moreover, we obtain our results using a real-world terrain, which results in a graph with 10,000 nodes and 80,000 links. Our results indicate that the developed algorithm outperforms the conventional Dijkstra algorithm for energy-cost minimization in terms of the length of the paths obtained, with a negligible excessive energy penalty.