An Open-Source Modular Benchmark for Diffusion-Based Motion Planning in Closed-Loop Autonomous Driving
Li Yun, Simon Thompson, Yidu Zhang, Ehsan Javanmardi, Manabu Tsukada · 2026
Read An Open-Source Modular Benchmark for Diffusion-Based Motion Planning in Closed-Loop Autonomous Driving by Li Yun, Simon Thompson, Yidu Zhang, Ehsan Javanmardi, Manabu Tsukada (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-01023.txt.
Description
Li and four colleagues build an open-source modular benchmark for diffusion-based motion planning inside real autonomous-driving stacks, not just standalone simulators. They split a monolithic planner of over eighteen thousand nodes into three runnable modules. They rebuild its solver loop in native C++, enabling runtime tuning and per-step inspection within ROS 2 and Autoware. Encoder caching cuts latency more than threefold, and second-order solving cuts final displacement error by 41 percent at three steps.
Catalog record
| CREATOR | Li Yun, Simon Thompson, Yidu Zhang, Ehsan Javanmardi, Manabu Tsukada |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31212 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.01023