Prediction-Driven Motion Planning: Route Integration Strategies in Attention-Based Prediction Models

TL;DR

This paper enhances attention-based motion prediction models by integrating navigation goals, improving the synergy between prediction and planning for automated vehicles, demonstrated through architectural strategies evaluated on the nuPlan dataset.

Contribution

It introduces novel methods for incorporating navigation information into attention-based prediction models, bridging the gap between multi-agent prediction and goal-oriented motion planning.

Findings

Navigation integration improves prediction accuracy.

Enhanced models better support goal-directed planning.

Results validated on the nuPlan dataset.

Abstract

Combining motion prediction and motion planning offers a promising framework for enhancing interactions between automated vehicles and other traffic participants. However, this introduces challenges in conditioning predictions on navigation goals and ensuring stable, kinematically feasible trajectories. Addressing the former challenge, this paper investigates the extension of attention-based motion prediction models with navigation information. By integrating the ego vehicle's intended route and goal pose into the model architecture, we bridge the gap between multi-agent motion prediction and goal-based motion planning. We propose and evaluate several architectural navigation integration strategies to our model on the nuPlan dataset. Our results demonstrate the potential of prediction-driven motion planning, highlighting how navigation information can enhance both prediction and planning tasks. Our implementation is at: https://github.com/KIT-MRT/future-motion.