Development of Safety Measures of Bicycle Trafflc by Observation wffh Deep-Leamlng, Drive Recorder Data, Probe Blcycle wlth LIDAR, and Connected Simulators
DOI: 10.25368/2022.476
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 outlines a four-year research project initiated in 2020 and funded by Japan’s Ministry of Land, Infrastructure, Transport and Tourism (MLIT) to develop and evaluate safety measures for bicycle traffic. The study is motivated by Japan’s high bicycle modal share (12%) coupled with relatively high bicycle-related fatalities compared to other nations in the IRTAD database. Specifically, the research targets serious collision types identified in government accident analysis, such as right-hook crashes at signalized intersections, left-turn vehicle collisions, and rear-end collisions. The project aims to address these risks, particularly in environments where bicycles share arterial roads with electric personal mobility vehicles, by establishing a comprehensive framework for evaluating proactive safety interventions. The methodology employs a multi-faceted approach combining real-world data collection with advanced simulation technologies. First, the researchers utilize deep learning techniques, specifically the Faster R-CNN method, to analyze image data and extract trajectories of left-turning vehicles and bicycles. This trajectory data is integrated with drive recorder data from occupational drivers and statistical accident data to calculate potential collision risks and define crash scenarios. Second, the study deploys a "probe bicycle" equipped with GPS, video cameras, and a Velodyne Lidar VLP-16 sensor. This device captures three-dimensional point data of surrounding objects within 100 meters, allowing for the detailed analysis of traffic conditions from the cyclist’s perspective and the verification of simulator accuracy against real-world trajectories and relative distances. Third, the project develops a connected simulator system to evaluate cooperative driving behaviors in virtual environments. Two simulator types are utilized: an 180-degree cylindrical screen system with additional monitors for rear and side views, and a head-mounted display providing a 360-degree field of view. These systems connect driving and cycling simulators, enabling users to interact within the same virtual space alongside computer-operated vehicles. This setup allows for the testing of specific interventions, such as varying corner components of protected intersections, to determine which conditions effectively ensure safety. The experimental design focuses on clarifying conditions for evaluating scenarios with and without interventions to propose effective crash prevention measures. The significance of this work lies in its integration of state-of-the-art observation tools and simulation environments to create a rigorous evaluation framework for bicycle safety. By combining deep-learning-based trajectory extraction, LiDAR-equipped probe data, and connected simulators, the project provides a robust method for identifying near-miss factors and testing cooperative driving behaviors. The paper serves as an overview of these developed tools and the study’s structure, noting that specific progress results and findings from the experiments will be presented separately at the conference. This approach contributes to the field by offering a systematic way to validate safety measures before their implementation in real-world traffic environments.
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. Discovered via author_sweep_intake on 2026-05-28.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | author_sweep | — | — | 2 | 2026-05-28 |
| archive | success | canonical_url | — | — | 1 | 2026-06-04 |
| extract | success | cached | — | — | 3 | 2026-06-15 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | — | — | — | 1 | 2026-05-28 |
| promote | success | — | — | — | 1 | 2026-06-04 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-15 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | success | — | — | — | 1 | 2026-06-15 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-15; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
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).
- Empirical Findings: crash risk outcomes
- Methodological Resource: tool software, dataset resource