Search based planners such as A* and Dijkstra's algorithm are proven methods for guiding today's robotic systems. Although such planners are typically based upon a coarse approximation of reality, they are nonetheless valuable due to their ability to reason about the future, and to generalize to previously unseen scenarios. However, encoding the desired… (More)
Fig. 4. An example marble maze and features used for planning. All features are computed by convolving, or “blurring”, discretized board geometry with two different kernel sizes (only one is shown here for each feature).