An Exploration of the Use of Artificial Intelligence for Enhanced Traffic Management, Operations and Safety
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 technical report, produced by the University of Texas at Austin’s Center for Transportation Research for the Texas Department of Transportation (TxDOT), investigates the application of artificial intelligence (AI) and machine learning (ML) to enhance traffic management, operations, and safety. The study was motivated by the need to understand the current state of AI in transportation, identify high-priority use cases for TxDOT, and assess the practical value, challenges, and limitations of implementing ML models. The research employed a multi-pronged approach comprising a comprehensive literature review, a stakeholder workshop, a data survey, the development of three prototype ML models for four prioritized use cases, and the field testing of one specific model. The methodology involved leveraging both emerging and traditional data sources. For safety analysis, the researchers used Wejo event data and Crash-Records Information System (CRIS) data to build supervised and unsupervised learning models. These models analyzed safety hot-spots and evaluated the impact of the pandemic on traffic patterns. To explore traffic signal timing, a microsimulation environment was utilized to test reinforcement learning models for real-time adjustment of signal timing plans in a frontage road setting. For travel time prediction, probe-based speeds from INRIX were combined with traffic volume data from TxDOT’s Intelligent Transportation Systems (ITS). Short-term travel time prediction models were selected for field testing due to their promising prototype results and the widespread availability of required data. Key findings from the prototype development demonstrated that ML models could effectively identify safety clusters and predict travel patterns. Specifically, preliminary ML-based travel time predictions were found to be 40% more accurate during peak periods compared to traditional approaches. Field testing involved training additional models in Austin and El Paso and deploying them via a web-based application for real-time evaluation. The performance evaluation focused on the ability to correctly identify the fastest route among competing alternatives. Results indicated that predictive ML models identified the correct fastest route more often than traditional naive methods. The study also developed a framework to expedite model training, testing, and evaluation to support real-time deployment, highlighting the feasibility of integrating these tools into existing transportation infrastructure. The significance of this work lies in its demonstration of the tangible benefits of AI in transportation operations, particularly in improving the accuracy of traveler information and optimizing traffic flow. The report provides a strategic approach to data acquisition and outlines pathways for implementation, including necessary hardware, software, and personnel considerations. By validating the superiority of ML over traditional methods in specific operational contexts, the study supports the broader adoption of AI technologies within state transportation departments. It offers a practical blueprint for leveraging diverse data sources, such as connected vehicle data and probe-based speeds, to enhance corridor management and safety, thereby contributing to more efficient and responsive transportation systems.
Key finding
Machine learning models for short-term travel time prediction were 40% more accurate than traditional approaches during peak periods and more reliably identified the fastest route among alternatives.
Methodology
mixed_methods
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: dataset resource
- Theoretical Contribution: computational model