A Statistical and Machine Learning Approach to Assess Contextual Complexity of the Driving Environment Using Autonomous Vehicle Data— Technology Transfer Activities
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 document outlines the technology transfer activities and outcomes of a research project titled "A Statistical and Machine Learning Approach to Assess Contextual Complexity of the Driving Environment Using Autonomous Vehicle Data." The study addresses the need for standardized, objective methods to evaluate the contextual complexity of driving environments, specifically to support the on-road assessment of medically at-risk drivers. The motivation stems from the limitations of current clinical evaluations, which often lack consistent metrics for dynamic roadway conditions. By leveraging transportation engineering data, the research aims to enhance the validity and reliability of driving assessments conducted by clinicians and occupational therapists. The methodology utilized open-source data from Waymo autonomous vehicles to capture dynamic traffic conditions. Researchers measured specific dynamic characteristics, including object density, proximity, and velocity, to quantify the complexity of the driving environment. The project employed statistical and machine learning approaches, specifically unsupervised clustering, to develop models capable of assessing contextual risk. These models were designed to classify routes based on their dynamic variables, creating a data-driven process for evaluating roadway contexts. The research culminated in a Ph.D. dissertation by Vijay Bendigeri, which focused on using safety performance models and autonomous vehicle data to establish standardized criteria for on-road evaluations. The primary outputs of the study include the completion of the aforementioned dissertation, a conference paper presented at the ASCE International Conference on Computing in Civil Engineering in 2021, and a student poster award from the Clemson IEEE ITSS student chapter. Additionally, the researchers conducted a seminar at the 46th annual conference of The Association for Driver Rehabilitation Specialists. This seminar educated clinicians on leveraging emerging technologies to improve the consistency of on-road driving tests. The findings demonstrated that autonomous vehicle data could effectively measure and classify the contextual complexity of routes, providing a tool for Driving Rehabilitation Specialists (DRSs) to score dynamic complexity during training and testing. The significance of this work lies in its potential to revolutionize the assessment of medically at-risk drivers. By providing objective tools to measure roadway context, the research bridges the gap between transportation engineering and the medical community. This collaboration allows clinicians to ensure that critical roadway components are considered during evaluations, thereby enhancing patient safety. The methodology offers broader implications for safety research, driver education, auto-insurance risk assessment, and autonomous vehicle route planning. Ultimately, the project establishes a foundational framework for using data-driven processes to improve the lives and safety of at-risk patients through more rigorous and standardized driving assessments.
Key finding
The development of data-driven models using autonomous vehicle data provides a standardized method for quantifying roadway contextual complexity to improve the validity of on-road driving assessments for medically at-risk drivers.
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 (39 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 | 37 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 19 | 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, tool software
- Theoretical Contribution: computational model