Trajectory Planning of Automated Vehicles Using Real-Time Map Updates

Szántó, Mátyás; Hidalgo, Carlos; González, Leonardo; Rastelli, Joshué Pérez; Asua, Estibaliz; Vajta, László · 2023 · Crossref

DOI: 10.1109/access.2023.3291350

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 limitation of static maps in Connected and Automated Vehicles (CAVs), which often suffer from high latency and low refresh frequencies, hindering robust trajectory planning in dynamic environments. The authors propose a novel map management framework that leverages Vehicle-to-Network (V2N) communication to provide real-time updates on dynamic obstacles. By aggregating sensor data from multiple vehicles via a crowdsourcing-like approach, the system aims to extend vehicle perception beyond onboard sensors, thereby improving safety and maneuvering efficiency, particularly for tasks like lane changes. The proposed solution utilizes a Candidate/Employed Map (C/EM) model integrated with the Message Queuing Telemetry Transport (MQTT) protocol. In this architecture, individual CAVs publish pre-processed obstacle observations from their LIDAR sensors to a central MQTT broker. A map management module subscribes to these updates, processes the data, and publishes the refined, real-time map back to the network. The system matches new observations against existing map data within a defined observable radius (default 50 meters) to update the central database. This framework was implemented within a six-block automated driving architecture, specifically integrating with the decision and control modules. The authors tested the system using a scenario involving both real and simulated CAVs performing lane change and braking maneuvers. The experimental results demonstrate that the simulated vehicle successfully optimized its trajectory planning using the real-time map updates received via the MQTT network. Crucially, the vehicle was able to detect and react to obstacles that were outside the range of its own onboard sensors, relying instead on the shared information from the map-management module. The system maintained the capability to perform automated maneuvers safely and effectively. The study confirms that the C/EM model facilitates low-latency map updating and allows vehicles to switch between maneuvers based on the dynamically updated central maps. The significance of this work lies in its demonstration of a viable, low-latency crowdsourced mapping solution for CAVs using standard IoT communication protocols. By shifting from static, infrequently updated maps to a real-time, collaborative mapping system, the approach enhances situational awareness and trajectory planning robustness. The paper establishes that leveraging V2N communication for dynamic obstacle mapping is a practical alternative to relying solely on onboard sensors or specialized infrastructure hardware, offering a scalable method for improving automated vehicle safety and efficiency in complex traffic 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.

• situational awareness