Bouzidi, Mohamed-Khalil (Free University of Berlin, Continental AG), Schlauch, Christian (Karlsruhe Institute of Technology), Scheuerer, Nicole (Continental Automotive Technologies), Yao, Yue (Continental Automotive GmbH), Klein, Nadja (Karlsruhe Institute of Technology), Goehring, Daniel (Freie Universität Berlin), Reichardt, Joerg (Continental Automotive Technologies GmbH)

Closing the Loop: Motion Prediction Models Beyond Open-Loop Benchmarks

Scheduled for presentation during the Regular Session "S41b-Motion Planning, Trajectory Optimization, and Control for Autonomous Vehicles" (FR-EA-T41), Friday, November 21, 2025, 14:10−14:30, Broadbeach 1&2

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on October 18, 2025

Abstract

Fueled by motion prediction competitions and benchmarks, recent years have seen the emergence of increasingly large learning-based prediction models, many with millions of parameters, focused on improving open-loop prediction accuracy by mere centimeters. However, these benchmarks fail to assess whether such improvements translate to better performance when integrated into an autonomous driving stack. In this work, we systematically evaluate the interplay between state-of-the-art motion predictors and trajectory planners. Our results show that higher open-loop accuracy does not always correlate with better closed-loop driving behavior and that other factors, such as temporal consistency of predictions and planner compatibility, also play a critical role. Furthermore, we investigate downsized variants of these models, and, surprisingly, find that in some cases models with up to 86% fewer parameters yield comparable or even superior closed-loop driving performance. Our code is available at https://github.com/continental/pred2plan.