Coupling SUMO with a Motion Planning Framework for Automated Vehicles

Klischat, Moritz; Dragoi, Octav; Eissa, Mostafa Essam; Althoff, Matthias · 2019 · OpenAlex-citations

DOI: 10.29007/1p2d

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the need for realistic simulation environments to test motion planning algorithms for automated vehicles. While virtual testing accelerates development compared to real-world drives, existing simulators often rely on predefined scenarios or lack the ability to simulate automated vehicles interacting with dynamic traffic. The authors propose coupling the open-source microscopic traffic simulator SUMO with their motion planning framework, CommonRoad, to enable closed-loop simulations where traffic participants react to host vehicles controlled by motion planners. The methodology involves creating an interface that exchanges trajectory data between SUMO and CommonRoad using SUMO’s TraCI API. A key challenge addressed is the synchronization of lane changes, as SUMO’s default implementation does not notify surrounding vehicles of a host vehicle’s lane change intent when positions are manually overridden, leading to unrealistic traffic behavior. To resolve this, the authors implement a synchronization algorithm that uses SUMO’s `changeLane` command during lane changes, allowing SUMO’s vehicle models to react appropriately. This approach accepts a small deviation between the planned and simulated trajectories to ensure realistic interactions. Additionally, the authors extend SUMO’s lane-changing models to replace constant lateral velocity with continuous acceleration, ensuring physically feasible lateral dynamics. Map conversion is handled by translating SUMO’s edge-and-junction format into CommonRoad’s lanelet format via OpenDRIVE. The results are demonstrated through a highway scenario involving a host vehicle and surrounding traffic. When using the standard position-overriding method (`moveToXY`), a surrounding vehicle failed to recognize the host vehicle’s lane change intent and attempted to overtake, forcing the host vehicle to abort its maneuver to avoid a collision. In contrast, using the proposed `changeLane` synchronization allowed the surrounding vehicle to recognize the intent early and yield, enabling the host vehicle to complete the lane change safely. The extended lane-changing model also produced smoother lateral movements compared to the original implementation. The significance of this work lies in providing a lightweight, open-source interface that enables realistic testing of motion planners against interactive traffic. By improving the realism of lateral dynamics and traffic reactions, the framework reduces false positives in testing and allows for more accurate evaluation of automated driving algorithms. The authors plan to integrate this interface into the CommonRoad benchmark suite to expand testing capabilities beyond predefined trajectories to include interactive decision-making scenarios.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• passenger motion sickness comfort

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Methodological Resource: tool software
• Theoretical Contribution: computational model