Real-Time Safety Diagnosis System for Connected Vehicles With Parallel Computing Architecture
archive: archived pipeline: cataloged verified
Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)
Summary
This report details the development of a real-time safety diagnosis system for connected vehicles (CVs), transitioning from a sequential computing paradigm to a parallel computing architecture. The work builds upon the previous STRIDE F4 project, which established a computational pipeline for diagnosing near-crash events using Basic Safety Messages (BSMs). The primary motivation for this upgrade was the increasing market penetration of CVs and the resulting surge in data volume, which necessitated faster processing speeds to ensure real-time safety warnings. The system defines a near-crash event as a situation involving both a vehicle conflict and at least one driver exhibiting abnormal driving status. The research focused on migrating the Driving Anomaly Detection (DAD) component from the cloud to In-Vehicle Computers (IVCs) to align with real-world operational constraints and leverage advancements in IVC hardware. The authors adopted a Domain-Specific Design (DSD) approach, configuring the system across three levels of abstraction: chip architecture, programming language, and parallelism module. For the Conflict Identification Model (CIM), the team recommended using ARM architecture with the C programming language, leveraging the chip’s built-in parallelism. For the DAD, they advocated for ARM architecture using Python on the CPU with multiprocessing for parallel computing. The system architecture was supported by a MySQL database for data handling, with parallel computing implemented via OpenCL in the cloud subsystem and OpenMP in the in-vehicle subsystem. Experimental evaluations were conducted using BSM data from connected vehicle pilot studies and crash data from the SHARPII naturalistic driving study. Tests were performed on Windows and MacOS platforms, including the Apple M1 chip, comparing performance across C, Python, and OpenMP. The study also assessed various Machine Learning packages for Object Detection (OD) to identify driving anomalies. Results indicated that standard major ML packages for OD failed to meet the system’s specific requirements for anomaly detection in this context. The parallel computing implementation successfully expedited data processing, demonstrating that the proposed DSD configuration effectively balances performance and productivity for in-vehicle applications. The significance of this work lies in its provision of a scalable, parallelized framework for real-time traffic safety diagnosis. By fully migrating anomaly detection to the vehicle edge and optimizing hardware-software configurations, the system addresses the latency and throughput challenges associated with transportation Big Data. The findings offer specific recommendations for configuring IVCs in future Intelligent Transportation Systems, highlighting the importance of DSD in overcoming the limitations of general-purpose parallel computing libraries. This approach ensures that safety diagnosis systems can keep pace with the growing complexity and volume of data generated by modern connected vehicle environments.
Key finding
The transition to a parallel computing architecture on in-vehicle computers using ARM architecture and specific programming languages enables real-time safety diagnosis, whereas standard machine learning object detection packages failed to meet system requirements.
Methodology
modeling
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 bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | rosap | — | — | 2 | 2026-05-23 |
| archive | success | — | — | — | 1 | 2026-05-23 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | success | — | — | — | 1 | 2026-05-23 |
| promote | success | — | — | — | 1 | 2026-05-23 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 3 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 24 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; 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).
- Methodological Resource: tool software