# Network Screening on Low-Volume Roads Using Risk Factors

^{1}

^{2}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Background

#### 2.1. Network Screening Methods Implemented in Practice

#### 2.2. Methods Proposed in the Literature

## 3. Motivation

## 4. Study Data

## 5. Methodology

- The EB expected number of total crashes was selected as a basis for network screening in the proposed method. This is to ensure effective network screening given the favorable performance of the EB method for LVRs [13,14]. In addition, multiple studies [15,16,17] have also reported favorable performance of the EB method over other methods used for network screening.
- The proposed method employs classified variables that can easily be compiled by local agencies using staff with limited technical skills. The risk factors were selected based on the findings from a previous study [9], that identified different risk factors for Oregon LVRs.

#### 5.1. Overview of the EB Method

#### 5.2. Classification of Roadway Variables

#### 5.3. Development of Regression Model

## 6. Study Results

#### 6.1. Level of Risk

#### 6.2. Model Variables

#### 6.3. Proposed Models

#### 6.4. Validation Results

## 7. Summary and Concluding Remarks

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Conflicts of Interest

## References

- National Highway Traffic Safety Administration (NHTSA). Rural Urban Comparison of Traffic Fatalities; NHTSA: Washington, DC, USA, 2018.
- Federal Highway Administration. (FHWA). Manual on Uniform Traffic Control Devices; FHWA: Washington, DC, USA, 2009. Available online: http://mutcd.fhwa.dot.gov/ (accessed on 28 March 2019).
- Gross, F.; Eccles, K.; Nabors, D. Low-Volume Roads and Road Safety Audits—Lessons Learned. Transp. Res. Rec. J. Transp. Res. Board
**2011**, 2213, 37–45. [Google Scholar] [CrossRef] - Federal Highway Administration. (FHWA). Systemic Safety Project Selection Tool; FHWA: Washington, DC, USA, 2013. Available online: http://safety.fhwa.dot.gov/systemic/ (accessed on 25 March 2019).
- Minnesota Department of Transportation (MnDOT). Minnesota’s Systemic Approach Integrates Safety Performance into Investment Decisions for Local Roads. 2014. Available online: https://www.fhwa.dot.gov/innovation/everydaycounts/edc_4/pdf/case_study_mn_oct2014.pdf (accessed on 25 March 2018).
- North Dakota Department of Transportation. Local Road Safety Program, N.d. Available online: https://www.dot.nd.gov/sites/www/files/documents/VIEW_WesternRegion_Agency_TheodoreRooseveltNationalPark.pdf (accessed on 25 March 2018).
- Zhong, C.; Sisiopiku, V.P.; Ksaibati, K.; Zhong, T. Crash Prediction on Rural Roads. In Proceedings of the 3rd International Conference on Road Safety and Simulation, Purdue University, Indianapolis, IN, USA, 14–16 September 2011. [Google Scholar]
- Aguero-Valverdea, J.; Wub, K.K.; Donnell, E.T. A multivariate spatial crash frequency model for identifying sites with promise based on crash types. Accid. Anal. Prev.
**2016**, 87, 8–16. [Google Scholar] - Al-Kaisy, A.; Veneziano, D.; Ewan, L.; Hossain, F. Risk Factors Associated with High Potential for Serious Crashes: Final Report; Report No. FHWA-OR-RD-16-05; Oregon Department of Transportation: Salem, OR, USA, 2015.
- Al-Kaisy, A.; Raza, S. A Novel Network Screening Methodology for Rural Low-Volume Roads. J. Transp. Technol.
**2023**, 13, 599–614. [Google Scholar] [CrossRef] - Harris, D.; Durdin, P.; Brodieb, C.; Tate, F.; Gardener, R. A road safety risk prediction methodology for low volume rural roads. In Proceedings of the Australasian Road Safety Conference, Gold Coast, Australia, 14–16 October 2015. [Google Scholar]
- American Association of State Highway and Transportation Officials (AASHTO). Highway Safety Manual, 1st ed.; AASHTO: Washington, DC, USA, 2010. [Google Scholar]
- Cafiso, S.; Di Silvestro, G.; Persaud, B.; Ara Begum, M. Revisiting variability of dispersion parameter of safety performance for two-lane rural roads. Transp. Res. Rec.
**2010**, 2148, 38–46. [Google Scholar] [CrossRef] - Al-Kaisy, A.; Huda, K.T. Empirical Bayes application on low-volume roads: Oregon case study. J. Saf. Res.
**2022**, 80, 226–234. [Google Scholar] [CrossRef] [PubMed] - Elvik, R. The predictive validity of Empirical Bayes estimates of road safety. Accid. Anal. Prev.
**2008**, 40, 1964–1969. [Google Scholar] [CrossRef] [PubMed] - Montella, A. A comparative analysis of hotspot identification methods. Accid. Anal. Prev.
**2010**, 42, 571–581. [Google Scholar] [CrossRef] [PubMed] - Manepalli, U.R.R.; Bham, G.H. An evaluation of performance measures for hotspot identification. J. Transp. Saf. Secur.
**2016**, 8, 327–345. [Google Scholar] [CrossRef] - Šenk, P.; Ambros, J.; Pokorný, P.; Striegler, R. Use of Accident Prediction Models in Identifying Hazardous Road Locations. Trans. Transp. Sci.
**2012**, 5, 223–232. [Google Scholar] [CrossRef] - Abebe, M.T.; Belayneh, M.Z. Identifying and Ranking Dangerous Road Segments a Case of Hawassa-Shashemene-Bulbula Two-Lane Two-Way Rural Highway, Ethiopia. J. Transp. Technol.
**2018**, 8, 151–174. [Google Scholar] [CrossRef] - Dixon, K.; Monsere, C.; Xie, F.; Gladhill, K. Calibrating the Future Highway Safety Manual Predictive Methods for Oregon State Highways; Final Report, SPR 684OTREC-RR-12-02; Oregon Department of Transportation: Salem, OR, USA, 2012.
- De’ath, G.; Fabricius, K.E. Classification and regression trees: A powerful yet simple technique for ecological data analysis. Ecology
**2000**, 81, 3178–3192. [Google Scholar] [CrossRef]

Risk Factors | Approximate Ranges of Variables | Categories | Terms | Statistics/Frequencies |
---|---|---|---|---|

Lane Width (LW) | LW < 11 ft | 1 | Narrower | Frequency: 1964 |

LW ≥ 11 ft | 2 | Wider | Frequency: 14,550 | |

Shoulder Width (SW) | SW < 2 ft | 1 | Narrower | Frequency: 5508 |

SW ≥ 2 ft | 2 | Wider | Frequency: 11,006 | |

Degree of Horizontal Curvature (DC) | DC = 0° | 0 | Straight | Frequency: 10,038 |

DC < 9° | 1 | Mild | Frequency: 5322 | |

9° ≤ DC < 28° | 2 | Moderate | Frequency: 615 | |

DC ≥ 28° | 3 | Sharp | Frequency: 539 | |

Grade (G) | G < 4% | 0 | Mild | Frequency: 12,593 |

G ≥ 4% | 1 | Steep | Frequency: 3921 | |

Driveway Density (DD) (driveways per mile) | Exact Number | Minimum: 0 First Quartile: 0 Median: 0 Mean: 4.479 Third Quartile: 0 Maximum: 100 | ||

Side Slope (SS) | Steep | 1 | Steep | Frequency: 3363 |

Moderate | 2 | Moderate | Frequency: 10,871 | |

Flat | 3 | Flat | Frequency: 2280 | |

Fixed Objects (FO) | Many | 1 | Many | Frequency: 10,700 |

Some | 2 | Some | Frequency: 4458 | |

Few | 3 | Few | Frequency: 1356 | |

Volume (V) | Exact Volume | Minimum: 60 First Quartile: 260 Median: 430 Mean: 496.5 Third Quartile: 620 Maximum: 2500 | ||

Dependent Variable: EB Expected Number of Crashes | Minimum: 0.0083 First Quartile: 0.0376 Median: 0.0561 Mean: 0.0994 Third Quartile: 0.0821 Maximum: 5.1193 |

Variables | Coefficients | p-Value | Significance at 95 Percent Confidence Level |
---|---|---|---|

Lane Width (LW) | −0.88 | <2 × 10^{−6} | ✓ |

Shoulder Width (SW) | −0.34 | <2 × 10^{−6} | ✓ |

Driveway Density (DD) | 0.016 | <2 × 10^{−6} | ✓ |

Volume (V) | 0.001 | <2 × 10^{−6} | ✓ |

Degree of Curvature (DC) | 0.24 | <2 × 10^{−6} | ✓ |

Side Slope (SS) | −0.31 | <2 × 10^{−6} | ✓ |

Few Fixed Objects (FO) | −0.21 | <2 × 10^{−6} | ✓ |

Adjusted R-squared = 0.915 |

Variables | Coefficients | p-Value | Significance at 95 Percent Confidence Level |
---|---|---|---|

Lane Width (LW) | −0.53 | <2 × 10^{−6} | ✓ |

Shoulder Width (SW) | −0.46 | <2 × 10^{−6} | ✓ |

Driveway Density (DD) | 0.02 | <2 × 10^{−6} | ✓ |

Degree of Curvature (DC) | 0.27 | <2 × 10^{−6} | ✓ |

Flat Side Slope (SS) | −0.28 | <2 × 10^{−6} | ✓ |

Few Fixed Objects (FO) | −0.25 | <2 × 10^{−6} | ✓ |

Adjusted R-squared = 0.905 |

Training Data | Testing Data | |||
---|---|---|---|---|

Mean | Total | Mean | Total | |

Actual Expected Crashes | 0.087 | 1213 | 0.084 | 294 |

Model Estimate | 0.085 | 1192 | 0.086 | 301 |

No Vol. Model Estimate | 0.076 | 1057 | 0.077 | 268 |

MBE | −0.0015 | 0.0018 | ||

MBE (no volume) | −0.011 | −0.007 | ||

RMSE | 0.18 | 0.19 | ||

RMSE (no volume) | 0.17 | 0.18 |

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |

© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Huda, K.T.; Al-Kaisy, A.
Network Screening on Low-Volume Roads Using Risk Factors. *Future Transp.* **2024**, *4*, 257-269.
https://doi.org/10.3390/futuretransp4010013

**AMA Style**

Huda KT, Al-Kaisy A.
Network Screening on Low-Volume Roads Using Risk Factors. *Future Transportation*. 2024; 4(1):257-269.
https://doi.org/10.3390/futuretransp4010013

**Chicago/Turabian Style**

Huda, Kazi Tahsin, and Ahmed Al-Kaisy.
2024. "Network Screening on Low-Volume Roads Using Risk Factors" *Future Transportation* 4, no. 1: 257-269.
https://doi.org/10.3390/futuretransp4010013