Investigating Individual Driver Performance: Applying DEA on Simulator Data

Babaee, Seddigheh; Shen, Yongjun; Hermans, Elke; Wets, Geert; Brijs, Tom · 2014 · Crossref

DOI: 10.1007/978-3-642-54927-4_59

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the critical need for evaluating individual driver performance to improve road safety, noting that inappropriate driver behavior contributes to over 90% of crashes. While existing research often focuses on aggregate crash data, there is a gap in analyzing individual risk factors to support proactive education and safety countermeasures. The authors aim to distinguish high-performing drivers from underperformers by constructing a composite driving performance index using data from a fixed-based driving simulator. The methodology employs Data Envelopment Analysis (DEA), specifically a multiple-layer DEA-based composite index model, to evaluate 129 participants aged 18–54. Drivers navigated a simulated two-lane rural road featuring a complex left-oriented compound curve. Performance was measured using three key indicators—speed, acceleration, and lateral position—at eight specific points along the trajectory (before, during, and after the curve). Raw data were processed via hierarchical cluster analysis and assigned ordinal grades, with lower grades indicating better performance. The model incorporated weight restrictions to prioritize performance during the curve and ensure balanced contribution from all three behavioral indicators. The results yielded a composite index score for each driver, allowing for relative ranking. Drivers with an index score of 1 were identified as best performers, while those with scores greater than 1 were classified as underperformers. The study categorized drivers into five performance groups, revealing that 1.55% were high performers, while 13.95% were low performers. Comparative analysis of best versus worst performers showed that underperformers exhibited higher speeds, excessive longitudinal and lateral acceleration, and greater variability in lateral position, particularly when entering and exiting the curve. Sensitivity analysis confirmed the robustness of the rankings, identifying specific drivers as consistently top-performing across various indicator exclusions. The significance of this work lies in its application of DEA to individual driver evaluation, providing a quantitative basis for identifying inefficient driving behaviors. The findings suggest that the composite index can effectively pinpoint specific weaknesses, such as poor speed control or lane positioning, enabling targeted training interventions. The authors conclude that this approach offers a valuable tool for proactive safety measures, moving beyond reactive crash analysis to individualized performance assessment. Future research is recommended to refine the model by using raw data rather than graded clusters and to expand the scope to other road types and 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.

• cognitive capacity variation
• lane positioning
• induced exposure
• exposure measurement
• telematics crash prediction

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: validation psychometrics, tool software
• Theoretical Contribution: computational model