aUToPath: Unified Planning and Control for Autonomous Vehicles in Urban Environments Using Hybrid Lattice and Free-Space Search (2505.09475v1)

Abstract: This paper presents aUToPath, a unified online framework for global path-planning and control to address the challenge of autonomous navigation in cluttered urban environments. A key component of our framework is a novel hybrid planner that combines pre-computed lattice maps with dynamic free-space sampling to efficiently generate optimal driveable corridors in cluttered scenarios. Our system also features sequential convex programming (SCP)-based model predictive control (MPC) to refine the corridors into smooth, dynamically consistent trajectories. A single optimization problem is used to both generate a trajectory and its corresponding control commands; this addresses limitations of decoupled approaches by guaranteeing a safe and feasible path. Simulation results of the novel planner on randomly generated obstacle-rich scenarios demonstrate the success rate of a free-space Adaptively Informed Trees* (AIT*)-based planner, and runtimes comparable to a lattice-based planner. Real-world experiments of the full system on a Chevrolet Bolt EUV further validate performance in dense obstacle fields, demonstrating no violations of traffic, kinematic, or vehicle constraints, and a 100% success rate across eight trials.

Unified Planning and Control for Autonomous Vehicles

The paper presents aUToPath, an advanced framework for the unified planning and control of autonomous vehicles operating in urban environments. The framework accounts specifically for the complexities associated with navigating cluttered and dynamic urban spaces. The authors introduce a novel hybrid motion planner combining pre-computed lattice maps with adaptive free-space sampling techniques. This approach is designed to efficiently produce optimal corridors navigable by autonomous vehicles within dense scenarios, facilitating the generation of feasible trajectories by integrating planning and control into a unified optimization problem.

The hybrid planner designated within aUToPath addresses significant limitations commonly associated with lattice-only approaches, such as resolution completeness and sub-optimal path outcomes in dense environments. By integrating sampling-based methodologies like Adaptively Informed Trees* (AIT*), the system leverages both informed sampling and a bidirectional search strategy. This integration is instrumental in enhancing the asymptotic optimality of path planning while maintaining computational efficiency, thus presenting a robust solution to global planning issues.

The paper further details the corridor generation methodology, which refines the initial trajectory into safe corridors through sequential convex programming (SCP)-based model predictive control (MPC). This method guarantees dynamically feasible, smooth paths that adhere to both the trajectory constraints and the real-time operational requirements expected in urban vehicular movement. Significant numerical results demonstrate the planner's capacity to navigate obstacle-rich terrains through simulations and validate its effectiveness via real-world experiments involving a Chevrolet Bolt EUV. Experiments displayed a 100% success rate across trials with no violations of vehicular, kinematic, or traffic constraints.

Practical implications of aUToPath lie in its ability to improve the reliability and safety of autonomous navigation in urban settings, potentially amplifying the capabilities of autonomous vehicles within the SAE AutoDrive Challenge framework and similar applications. The approach minimizes computational load while integrating environmental data, thus promoting real-time adaptability in varying traffic situations without compromising safety.

Future developments may explore enhanced robustness against upstream errors, such as false detections or localization drift, and could consider augmenting the architecture with machine learning techniques to further refine predictability and enhance adaptive control behaviors. The theoretical implications suggest a definitive move towards integrated systems that resolve both path planning and control in an optimized, streamlined process, paving the way for more efficient and reliable autonomous vehicle deployments in complex environments.