Anthony Dontoh

Research & Projects

An improved ResNet50 model for predicting pavement condition index (PCI) directly from pavement images

Andrews Danyo, Anthony Dontoh, Armstrong Aboah

Road Materials and Pavement Design, 2025 — Taylor & Francis

Taylor & Francis DOI

Abstract: Accurately predicting the Pavement Condition Index (PCI), a measure of roadway conditions, from pavement images is crucial for infrastructure maintenance. This study proposes an enhanced version of the Residual Network (ResNet50) architecture, integrated with a Convolutional Block Attention Module (CBAM), to predict PCI directly from pavement images without additional annotations. By incorporating CBAM, the model autonomously prioritises critical features within the images, improving prediction accuracy. Compared to the original baseline ResNet50 and DenseNet161 architectures, the enhanced ResNet50-CBAM model achieved a significantly lower mean absolute percentage error (MAPE) of 58.16%, compared to the baseline models that achieved 70.76% and 65.48% respectively. These results highlight the potential of using attention mechanisms to refine feature extraction, ultimately enabling more accurate and efficient assessments of pavement conditions. This study emphasises the significance of targeted feature refinement in advancing automated pavement analysis through attention mechanisms.
```
@article{Danyo30042025,
  author   = {Andrews Danyo and Anthony Dontoh and Armstrong Aboah},
  title    = {An improved ResNet50 model for predicting pavement condition index (PCI) directly from pavement images},
  journal  = {Road Materials and Pavement Design},
  volume   = {0},
  number   = {0},
  pages    = {1--18},
  year     = {2025},
  publisher= {Taylor \& Francis},
  doi      = {10.1080/14680629.2025.2498632},
  url      = {https://doi.org/10.1080/14680629.2025.2498632},
  eprint   = {https://doi.org/10.1080/14680629.2025.2498632}
}
```
Rail’s Workforce Frontier: Building America’s Talent Pipeline

Anthony Dontoh, Stephanie Ivey, Janey Camp, Kassandra Arellano, Stefano Alessandro Blasoni, Dustie Flowers, Jaila Kimbro, Logan Nelson Sirbaugh, Martin Lipinski, Tanisha Hall, Kristen Jud, Jim McAteer, Catherine Armwood-Gordon, Shala Blue

Transportation Research Record (TRR), 2025 — SAGE

TRR DOI

Abstract: Women in the rail industry hold only about 10% of jobs, and even fewer in specific technical roles. The underrepresentation has the potential to hamper innovation and productivity while limiting full utilization of diverse talent perspectives. With an aging transportation workforce facing imminent retirements, failing to recruit and retain underrepresented groups, particularly women, threatens long-term workforce sustainability and service continuity. Teams with greater representation of women are shown to deliver superior organizational performance through enhanced decision-making and problem-solving capabilities. In rail operations, where complex scheduling, safety compliance, and operational efficiency are critical, diverse teams can provide more adaptive and creative approaches to these essential challenges. This study aims to offer comprehensive insights into the history of the participation of women in rail, barriers to balanced participation, and strategies to enhance women’s representation in the rail industry. Employing an extensive literature review, case study analyses, identification of professional organizations focused on the representation of women in technical fields, the findings highlight that although women’s participation remains limited, there is noticeable progress. Success stories from other countries and male-dominated industries provide valuable strategies for attracting and retaining women in rail careers. The study identifies entrenched industry perceptions as significant barriers. It concludes that an industry-wide, comprehensive effort is important to shift these perceptions and increase women’s participation in the rail sector.
```
@article{doi:10.1177/03611981251352520,
author = {Anthony Dontoh and Stephanie Ivey and Janey Camp and Kassandra Arellano and Stefano Alessandro Blasoni and Dustie Flowers and Jaila Kimbro and Logan Nelson Sirbaugh and Martin Lipinski and Tanisha Hall and Kristen Jud and Jim McAteer and Catherine Armwood-Gordon and Shala Blue},
title ={Rail’s Workforce Frontier: Building America’s Talent Pipeline},
journal = {Transportation Research Record},
volume = {0},
number = {0},
pages = {03611981251352520},
year = {0},
doi = {10.1177/03611981251352520},
URL = {https://doi.org/10.1177/03611981251352520},
eprint = {https://doi.org/10.1177/03611981251352520}
}
```

Advancing Gender Equity in the DOT Workforce

TRB & National Academies; Editors: Stephanie Ivey, Susan Gallagher, Anthony Dontoh, Sarah Hashemikamangar, Dustie Flowers, Kassandra Arellano, Prashant Jha, Logan Sirbaugh, Jaila Kimbro

NCHRP Synthesis 618 — National Academies Press, 2024

NASEM DOI

Abstract: Synthesis documenting current DOT practices for advancing gender equity; summarizes trends, barriers, and actionable strategies across agencies.

@BOOK{NAP27646,
  author    = "Transportation Research Board and National Academies of Sciences, Engineering, and Medicine",
  editor    = "Stephanie Ivey and Susan Gallagher and Anthony Dontoh and Sarah Hashemikamangar and Dustie Flowers and Kassandra Arellano and Prashant Jha and Logan Sirbaugh and Jaila Kimbro",
  title     = "Advancing Gender Equity in the DOT Workforce",
  isbn      = "978-0-309-70950-7",
  doi       = "10.17226/27646",
  url       = "https://nap.nationalacademies.org/catalog/27646/advancing-gender-equity-in-the-dot-workforce",
  year      = 2024,
  publisher = "The National Academies Press",
  address   = "Washington, DC"
}

Identifying Critical Knowledge Gaps and Assessing Organizational Readiness for Improved Knowledge Management

Stephanie Ivey, Brian Janz, Janey Camp, Anthony Dontoh, Sarah Hashemikamangar, Kassandra Arellano, Logan Sirbaugh, Stefano Blasoni, Dustie Flowers, Charlotte Rector, Drew Davis, Catherine Armwood-Gordon, Kimberly Triplett, Imani Rivers

TDOT Final Report RES2023-01 — University of Memphis, 2024

TDOT Report

Abstract: For an organization to develop a sustainable strategy that supports technical capacity, workforce retention, and innovation, it is essential to develop a strong framework for knowledge management (KM) both within individual divisions as well as organization wide. For state departments of transportation (DOTs), KM has become a topic of increasing urgency as the aging and knowledgeable workforce that has been the subject of recent concern has now begun large-scale retirements. Many DOTs are faced with significant threats to organizational knowledge and technical capacity as knowledge holders leave and new workers are onboarded without the infrastructure for effective knowledge transfer. This project examined TDOT’s knowledge assets, current KM practices, and organizational culture with respect to KM. This research gaps in KM practice and a set of recommendations to support TDOT in institutionalizing a robust practice and culture of KM and establishing a continuous assessment process for the organization to evolve in its KM maturity. It is expected that the results of this research will enable TDOT to develop a robust and sustainable process for the efficient transfer of knowledge that improves worker retention, technical capacity, KM culture, and continued innovation resulting in successful fulfillment of its mission.
```
Title: Identifying Critical Knowledge Gaps and Assessing Organizational Readiness for Improved Knowledge Management
Creators: Ivey, Stephanie; Janz, Brian; Camp, Janey; Dontoh, Anthony; Hashemikamangar, Sarah; Arellano, Kassandra; Sirbaugh, Logan; Blasoni, Stefano; Flowers, Dustie; Rector, Charlotte; Davis, Drew; Armwood-Gordon, Catherine; Triplett, Kimberly; Rivers, Imani
Corporate Creator: University of Memphis
Corporate Contributors: Tennessee Department of Transportation (Long Range Planning Research Office); FHWA
Date: 2024-10-01
Report Number: RES2023-01
URL: https://rosap.ntl.bts.gov/view/dot/78009
```
Temporal Transferability of Regional Trip Generation Models Updated with Spatially Transferred Data

Anthony Dontoh

Tennessee Technological University, 2022 — ProQuest

ProQuest

Abstract: Federal law in the United States mandates that states and metropolitan regions develop Long Range Transportation Plans that detail their major infrastructure needs that would allow for safe and efficient travel in the future. Determination of facility needs requires the forecast of travel volumes and their comparison to the capacity of available transportation infrastructure. Transportation engineers typically accomplish the forecasting task using a sequential procedure called the four-step urban transportation modeling system. The procedure requires the collection of household and travel data in across-sectional household travel behavior survey. Due to the high cost of conducting these surveys, metro regions undertake them on average about every 10 to 20 years. During the long inter-survey times, the modeled travel behavior of a regional population could change in such a way that the forecasts generated by the models could have large errors in them. Metro regions in a state normally do not synchronize the conduct of these surveys. This raises the question of whether data collected in a travel survey undertaken in one metro region within a state can be updated with data collected in surveys undertaken in a subsequent year but in a different metro region within the same state for model development. This research seeks to investigate the above using data collected in Nashville, Knoxville, and Chattanooga. Specific objectives are: (1) to investigate the temporal transferability of trip generation models developed using data collected in Nashville in 1998 and 2012; (2) to investigate the temporal transferability of metropolitan trip generation models updated with subsequently collected data from another metropolitan region in the state of Tennessee; and (3) to investigate whether greater similarities in demographic characteristics between two metro regions translates into improved transfer performance. The analysis was conducted using multiple linear regression analysis and Bayesian statistics. The conclusions reached were (1) that the transfer performance of most of the Nashville 1998 trip-purpose models updated with subsequent data from another metro region improved in accuracy by a margin ranging from four to fifteen percent; (2) the method employed in model updating influences transfer performance with the Bayesian updating method giving the best improvement (the exception was home-based school trips); (3) greater similarities in demographics between two metro regions does not guarantee improved transfer performance; and (4) there are improvements in transfer performance when there is a smaller difference between an update survey data’s year and the application context year.
```
@mastersthesis{Dontoh2022TemporalTransferability,
  author  = {Anthony Dontoh},
  title   = {Temporal Transferability of Regional Trip Generation Models Updated with Spatially Transferred Data},
  school  = {Tennessee Technological University},
  address = {Cookeville, TN},
  year    = {2022},
  note    = {ProQuest Dissertations \& Theses Global},
  pages   = {106}
}
```

Anthony Dontoh

About

Open to Collaboration

News

Research & Teaching Interests

Research & Projects

Contact