Aerial assisted path planning for terrestrial rover without complete environment map

There are several existing methods available for planning paths for terrestrial rovers including graph searching, artificial potential fields, neural networks, fuzzy logic, genetic algorithms and ant colony optimization. Of these, the graph search technique is the most popular due to ease of implementation on rovers which don´t have a complete map of the area in which they are navigating. Regardless of the search technique, using a rover to plan out the path is expensive in terms of energy consumption and time. In a situation where the rover does not have a complete map of the environment it has no choice but to explore and discover obstacles to reach the final destination. The authors of this provide a solution to this problem by attaching a manually controlled drone to the rover. The drone will scout out the area that the robot has to travel in, identify obstacles through stereo- imaging, calculate the shortest path and instruct the rover to follow this optimized path. This approach reduces the need for various sensors on board for autonomous navigation, thus reducing power consumption of the rover and time taken for exploration, thus significantly increasing the energy efficiency.