Knowledge Acquisition Methods for the IHDS Diagnostic Review Expert System

Hughes, Warren E.; Schuckel, Clint; Smith, Michelle; McGee, Hugh W. · 1997 · ROSA P / United States. Department of Transportation. Federal Highway Administration

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 presents a feasibility study conducted for the Federal Highway Administration (FHWA) to determine the most effective methods for developing the knowledge base for the Diagnostic Review Component (DRC). The DRC is an expert system module within the Interactive Highway Safety Design Model (IHSDM), designed to help highway designers proactively evaluate the safety implications of geometric design alternatives, particularly for two-lane rural roads. The research was motivated by the lack of systematic procedures to assess safety trade-offs during the design process, where decisions are often made based on cost rather than empirical safety evidence. The study aimed to investigate alternative knowledge acquisition methods and develop an experimental plan for constructing the DRC’s knowledge base, rather than building the base itself. The researchers evaluated four primary methods for acquiring knowledge: interviews with state highway design personnel and safety experts, in-depth forensic investigations of specific accident sites, reviews of highway safety research and safety-audit literature, and analyses of Highway Safety Improvement Program (HSIP) data. The study involved interviewing experts in safety, geometric design, accident investigation, traffic engineering, and human factors. Additionally, multidisciplinary teams conducted field investigations at four specific rural intersection sites in Minnesota to assess the utility of forensic site analysis. The team also reviewed existing literature on problematic design situations and analyzed documentation from state HSIP projects to identify common safety improvements and design deficiencies. The findings detailed the advantages, disadvantages, and required effort for each method. Expert interviews provided broad insights but required careful screening to ensure relevance. Forensic site investigations offered detailed, context-specific data but were resource-intensive and difficult to generalize into design guidelines. Literature reviews yielded established design guidelines and identified problematic geometric combinations, while HSIP data provided practical examples of implemented safety improvements. The study synthesized these findings to propose a systematic process for developing an interactive checklist and an automated review function for the DRC. It highlighted critical issues such as defining problem emphasis areas, determining the required specificity level for rules, and managing the interaction between the DRC and the Policy Review Module. The significance of this work lies in its contribution to the development of the IHSDM, a tool intended to integrate safety explicitly into the highway design process. By identifying viable methods for knowledge acquisition, the report provides a roadmap for creating a robust expert system that can flag geometric deficiencies and suggest corrective actions before construction. The authors concluded with recommendations for subsequent research, including the development of a preliminary knowledge base, a prototype expert system, and the expansion of the model to incorporate general AASHTO design guidelines. This feasibility study establishes the foundational approach for translating complex safety research and expert knowledge into a usable, computer-based design tool.

Key finding

The study concludes that a combination of expert interviews, forensic site investigations, literature reviews, and HSIP data analysis provides a feasible foundation for constructing the DRC knowledge base, with specific recommendations for a prototype development plan.

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).

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.

• naturalistic crash near crash
• expert witness methodology
• rail grade crossings

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: dataset resource