Parting with Misconceptions about Learning-based Vehicle Motion Planning
This work addresses vehicle motion planning for autonomous driving, revealing that current learning-based approaches are limited and that simpler methods can achieve better performance, which is incremental as it builds on existing datasets and challenges.
The paper tackles the misalignment between short-term planning and long-horizon ego-forecasting in vehicle motion planning, finding that these tasks should be addressed independently and that simple rule-based priors like centerline selection outperform complex learning-based methods, leading to a winning entry in the nuPlan planning challenge 2023.
The release of nuPlan marks a new era in vehicle motion planning research, offering the first large-scale real-world dataset and evaluation schemes requiring both precise short-term planning and long-horizon ego-forecasting. Existing systems struggle to simultaneously meet both requirements. Indeed, we find that these tasks are fundamentally misaligned and should be addressed independently. We further assess the current state of closed-loop planning in the field, revealing the limitations of learning-based methods in complex real-world scenarios and the value of simple rule-based priors such as centerline selection through lane graph search algorithms. More surprisingly, for the open-loop sub-task, we observe that the best results are achieved when using only this centerline as scene context (i.e., ignoring all information regarding the map and other agents). Combining these insights, we propose an extremely simple and efficient planner which outperforms an extensive set of competitors, winning the nuPlan planning challenge 2023.