Open-Source, Cost-Aware Kinematically Feasible Planning for Mobile and Surface Robotics

TL;DR

Smac Planner addresses the challenge of globally planning kinematically feasible paths for non-circular and nonholonomic robots by introducing Cost-Aware variants of A*, Hybrid-A*, and State Lattice planners. Its templated A* core and modular, middleware-agnostic design enable rapid development and integration within ROS 2 Nav2, with implementations that maintain feasibility while improving performance in complex environments. Empirical results in simulation and a large warehouse demonstrate faster planning and near-optimal path quality compared to baselines, validating the practical impact for delivery, warehousing, and surface robotics. By providing open-source, high-performance planners and minimal-cost control set generation, Smac Planner lowers barriers to deploying sophisticated kinodynamic navigation across diverse platforms.

Abstract

We present Smac Planner, an openly available, search-based planning framework that addresses the critical need for kinematically feasible path planning across diverse robot platforms. Smac Planner provides high-performance implementations of Cost-Aware A*, Hybrid-A*, and State Lattice planners that can be deployed for Ackermann, legged, and other large non-circular robots. Our framework introduces novel "Cost-Aware" variations that significantly improve performance in complex environments common to mobile robotics while maintaining kinematic feasibility constraints. Integrated as the standard planning system within the popular ROS 2 Navigation stack, Nav2, Smac Planner now powers thousands of robots worldwide across academic research, commercial applications, and field deployments.

Figures (8)

• Figure 1: Examples of industrial and marine surface robots using Nav2 uniquely enabled by our Smac Planner open-source framework. More examples can be found at https://docs.nav2.org/about/robots.html.
• Figure 1: A set of example paths of the Smac Planner in a 33,600$m^2$ warehouse serviced by Locus Robotics.
• Figure 2: Outline of the A* and key node template methods.
• Figure 2: Cost-Aware 2D-A* with variable cost penalties.
• Figure 3: Illustrative example where higher costs are applied to a region to dissuade travel. This soft constraint may be introduced due to increased risk or danger in that area (red). On the left, the Cost-Aware Obstacle Heuristic uses this constraint and finds a longer path, going around the danger zone to achieve a lower path cost. On the right, the binary Obstacle Heuristic ignores this constraint and navigates directly through the region for a modestly lower path length. Our heuristic steered towards the goal and maintained acceptably navigable distances from inflated obstacles.